# neurostack - Doramagic AI Context Pack

> Positioning: a pre-install experience and judgment asset. It helps the host AI get off to a good start, but it does not mean the project has already been installed, run, or validated.

## Sufficiency Principle

- **Sufficiency over compression**: The AI Context Pack should be sufficient for the host AI to understand the project's value, capability boundaries, entrypoints, risks, and evidence sources before starting work; it may be layered, but it does not aim for the shortest possible summary.
- **Compression policy**: Compress only noise and duplication, never context that affects judgment or the quality of the work.

## How the Host AI Should Use This

You are reading the AI Context Pack that Doramagic compiled for neurostack. Treat it as pre-work context: help the user understand who it fits, what it can do, how to start, what must be verified after install, and where the risks are. Do not claim that you have already installed, run, or executed the target project.

## Claim Consumption Rules

- **Fact source**: Repo Evidence + Claim/Evidence Graph; the Human Wiki only supplies salience, terminology, and narrative structure.
- **Minimum status for a fact**: `supported`
- `supported`: May be used as a project fact, but the answer must cite the claim_id and evidence path.
- `weak`: Usable only as a low-confidence lead; the user must be asked to keep verifying.
- `inferred`: Usable only for risk notes or open questions; must not be packaged as a project fact.
- `unverified`: Must not be used as fact; state clearly that evidence is insufficient.
- `contradicted`: Must show the conflicting sources and must not force a single version on the user's behalf.

## Who It Fits Best

- **Users who want to understand an open-source project's value and boundaries before installing**: Current evidence comes mainly from project documentation. Evidence: `README.md` Claim: `clm_0002` supported 0.86

## What It Can Do

- **Command-Line Startup or Install Flow** (Verify after install): The project documentation contains runnable commands; real use requires running them in a local or host environment. Evidence: `CLAUDE.md` Claim: `clm_0001` supported 0.86

## How to Start

- `npm install -g neurostack    # bootstraps CLI, Python, uv, deps` Evidence: `CLAUDE.md` Claim: `clm_0003` supported 0.86
- `pip install neurostack[api]` Evidence: `CLAUDE.md` Claim: `clm_0004` supported 0.86

## Continue-or-Stop Decision Card

- **Current recommendation**: Needs admin / security approval
- **Why**: Continuing may involve secrets, accounts, external services, or sensitive context; get admin or security approval first.

### 30-Second Read

- **What to do now**: Needs admin / security approval
- **Minimum safe next step**: Run Prompt Preview first; if credentials or an enterprise environment are involved, get approval before trialing
- **Do not trust yet**: Role quality and task fit cannot be trusted directly.
- **Continuing will touch**: Role selection bias, Command execution, Host AI configuration

### What You Can Trust Now

- **Target-audience signal: Users who want to understand an open-source project's value and boundaries before installing** (supported): Backed by a supported claim or project evidence, but that still is not the same as real install results. Evidence: `README.md` Claim: `clm_0002` supported 0.86
- **Capability exists: Command-Line Startup or Install Flow** (supported): You can trust that the project contains signals of this capability; whether it fits your specific task still needs trial or after-install verification. Evidence: `CLAUDE.md` Claim: `clm_0001` supported 0.86
- **There are Quick Start / install-command signals** (supported): You can trust that the docs mention a startup or install entrypoint; do not run it directly in your primary environment because of that. Evidence: `CLAUDE.md` Claim: `clm_0003` supported 0.86

### What You Cannot Trust Yet

- **Role quality and task fit cannot be trusted directly.** (unverified): A role library proves there are many roles; it does not prove each one fits your specific task or that a role produces high-quality results.
- **Do not treat role copy as real execution capability.** (unverified): Before install you can only judge whether the role description and task profile match; you cannot prove it can complete the task inside the host AI.
- **Real output quality cannot be trusted before install.** (unverified): Prompt Preview can only show how it guides you; it cannot prove result quality in the real project.
- **Host AI version compatibility cannot be trusted before install.** (unverified): Host loading rules and version differences across Claude, Cursor, Codex, Gemini, and others must be verified in a real environment.
- **That it will not pollute your existing host AI's behavior cannot be trusted directly.** (inferred): Skill, plugin, and AGENTS/CLAUDE/GEMINI instructions may change the host AI's default behavior. Evidence: `CLAUDE.md`
- **Safe rollback cannot be assumed by default.** (unverified): Unless the project clearly provides uninstall and recovery instructions, verify in an isolated environment first.
- **After a real install, is it compatible with the user's current host AI version?** (unverified): Compatibility can only be verified in the actual host environment.
- **Does the project's output quality meet the user's specific task?** (unverified): The pre-install preview can only show flow and boundaries; it cannot replace real evaluation.

### What Continuing Will Touch

- **Role selection bias**: The user's judgment about which expert role should handle the task. Why: Picking the wrong role makes the AI answer from the wrong expert perspective, wasting time or misleading decisions.
- **Command execution**: Package managers, network downloads, the local plugin directory, project config, or the user's home directory. Why: Running the very first command can already change your environment; decide whether it is worth running first. Evidence: `CLAUDE.md`
- **Host AI configuration**: The plugin, Skill, or rule-loading config of hosts like Claude/Codex/Cursor/Gemini/OpenCode. Why: Host configuration changes how the AI works afterward and may conflict with the user's existing rules. Evidence: `CLAUDE.md`
- **Local environment or project files**: Install results, plugin caches, project config, or local dependency directories. Why: The write scope and rollback path cannot be proven before install and need isolated verification. Evidence: `CLAUDE.md`
- **Environment variables / API keys**: Project entry docs explicitly showing API key, token, secret, or account credential configuration. Why: If a real install needs credentials, use test credentials first and go through a permission/compliance review. Evidence: `CLAUDE.md`, `manifest.json`
- **Host AI context**: The AI Context Pack, Prompt Preview, Skill routing, risk rules, and project facts. Why: Importing context affects the host AI's later judgment, so avoid packaging unverified items as facts.

### Minimum Safe Next Steps

- **Run Prompt Preview first**: Use an interactive trial to verify the task profile and role match first; do not import the whole role library up front. (applies when: Applies to any project, especially when output quality is unknown.)
- **Trial-install only in an isolated directory or a test account**: Avoid letting install commands pollute your primary host AI, real projects, or home directory. (applies when: When there are signals of command execution, plugin config, or local writes.)
- **Back up your host AI configuration first**: Skill, plugin, and rule files may change the default behavior of Claude/Cursor/Codex. (applies when: When there is a plugin manifest, a Skill, or a host rule entrypoint.)
- **Do not use real production credentials**: Once an environment variable / API key enters the host or toolchain, it can create account and compliance risk. (applies when: When environment signals like API, TOKEN, KEY, or SECRET appear.)
- **After install, verify just one minimal task**: Verify loading, compatibility, output quality, and rollback first, then decide whether to use it deeply. (applies when: When moving from a trial into a real workflow.)

### Exit Plan

- **Preserve the pre-install state**: Record the original host config and project state so you can later judge whether it is recoverable.
- **Be ready to remove the host plugin / Skill / rule entrypoint**: If behavior is off after the trial install, you can restore the host AI to its pre-trial state.
- **Keep a record of the original role selection**: If output goes off-topic, you can return to the task-profiling stage and reselect a role instead of pushing on with the wrong one.
- **Record the install commands and written paths**: Without clear uninstall instructions, you at least need to know which directories or configs to clean up manually.
- **Be ready to revoke test API keys or tokens**: If test credentials leak or are misused, you can cut losses quickly.
- **If there is no rollback path, do not enter your primary environment**: No rollback is a blocker before continuing; do not proceed on trust or luck.

## What Can Only Be Previewed

- Explain who the project fits and what it can do
- Demonstrate a typical conversation flow based on project docs
- Help the user decide whether it is worth installing or researching further

## What Must Be Verified After Install

- Actually installing the Skill, plugin, or CLI
- Running scripts, modifying local files, or accessing external services
- Verifying real output quality, performance, and compatibility

## Boundary & Risk Decision Card

- **Mistaking the pre-install preview for a real run**: The user may overestimate how much configuration, permission, and compatibility verification the project has already done. Mitigation: Clearly separate prompt_preview_can_do from runtime_required. Claim: `clm_0005` inferred 0.45
- **Command execution will modify the local environment**: Install commands may write to the user's home directory, the host plugin directory, or project configuration. Mitigation: Run in an isolated environment or a test account first. Evidence: `CLAUDE.md` Claim: `clm_0006` supported 0.86
- **To confirm**: After a real install, is it compatible with the user's current host AI version?. Why: Compatibility can only be verified in the actual host environment.
- **To confirm**: Does the project's output quality meet the user's specific task?. Why: The pre-install preview can only show flow and boundaries; it cannot replace real evaluation.
- **To confirm**: Do the install commands require network access, permissions, or global writes?. Why: This affects install risk in both enterprise and personal environments.

## Pre-Work Working Context

### Loading Order

- First read how_to_use.host_ai_instruction to establish the boundaries of this pre-install judgment asset.
- Read claim_graph_summary to confirm facts come from the Claim/Evidence Graph, not the Human Wiki narrative.
- Then read intended_users, capabilities, and quick_start_candidates to judge whether the user is a match.
- When you need to carry out a concrete task, check role_skill_index first, then evidence_index.
- For real install, file modification, network access, performance, or compatibility questions, turn to risk_card and boundaries.runtime_required.

### Task Routes

- **Command-Line Startup or Install Flow**: State that this is an after-install capability first, then give a pre-install checklist. Boundary: Must be verified after a real install or run. Evidence: `CLAUDE.md` Claim: `clm_0001` supported 0.86

### Context Scale

- Total files: 156
- Important-file coverage: 40/156
- Evidence index entries: 78
- Role / Skill entries: 54

### Handling Insufficient Evidence

- **missing_evidence**: State that evidence is insufficient and ask the user for the target file, a README section, or after-install verification records; do not fill in facts.
- **out_of_scope_request**: State that the task is beyond the current AI Context Pack's evidence scope and suggest the user check the Human Manual or verify after a real install.
- **runtime_request**: Provide a pre-install checklist and command sources, but do not run commands for the user or claim they have been run.
- **source_conflict**: Show the conflicting sources side by side, mark them as unverified, and do not force a single version.

## Prompt Recipes

### Fit assessment

- Goal: Judge whether this project fits the user's current task.
- Expected output: A fit conclusion, key reasons, evidence citations, what can be previewed before install, what must be verified after install, and a next-step recommendation.

```text
Based on the AI Context Pack for neurostack, ask me 3 necessary questions first, then judge whether it fits my task. The answer must cover: who it fits, what it can do, what it cannot do, whether it is worth installing, and where the evidence comes from. Every project fact must cite evidence_refs, source_paths, or a claim_id.
```

### Pre-install experience

- Goal: Let the user feel the core workflow before installing, while avoiding packaging the preview as real capability or a marketing promise.
- Expected output: An experience script with boundary labels, an after-install verification checklist, and a cautious recommendation; with no real-run promises or strong marketing language.

```text
Treat neurostack as a pre-install experience asset, not an already-installed tool or a real runtime environment.

Output exactly four parts:
1. Ask me 3 necessary questions first.
2. Give an "experience script": use the three labels [Previewable before install], [Must verify after install], and [Insufficient evidence] to show how it might guide the workflow.
3. Give an after-install verification checklist: list which capabilities can only be confirmed after a real install, real host loading, and a real project run.
4. Give a cautious recommendation: only "worth researching/trialing further", "add information before deciding", or "not recommended to continue"; do not endorse the project.

Hard boundaries:
- Do not claim you have installed, run, executed tests, modified files, or produced real results.
- Do not write promise-like phrasing such as "auto-adapts", "guarantees passing", "perfect fit", or "strongly recommend installing".
- If you describe how it works after install, you must use a conditional such as "if installed successfully and the host loads the Skill correctly, it might...".
- The experience script may only be written as "example lines / hypothetical flow": use "might ask / might suggest / might show", not "has written, has generated, has passed, is running, is generating".
- Prompt Preview does not hand out install commands; if the user is ready to trial, only prompt them to read Quick Start and the Risk Card first and to verify in an isolated environment.
- Every project fact must come from a supported claim, evidence_refs, or source_paths; inferred/unverified items can only be risks or open questions.

```

### Role / Skill selection

- Goal: Pick the best-matching asset from the project's roles or Skills.
- Expected output: A list of candidate roles or Skills, each with an applicable scenario, evidence paths, risk boundary, and whether after-install verification is needed.

```text
Read role_skill_index and recommend 3-5 of the most relevant roles or Skills for my target task. For each recommendation, state the applicable scenario, likely output, risk boundary, and evidence_refs.
```

### Risk pre-check

- Goal: Identify environment, permission, rule-conflict, and quality risks before installing or adopting.
- Expected output: A checklist of environment, permission, dependency, license, host-conflict, quality risk, and unknown items.

```text
Based on risk_card, boundaries, and quick_start_candidates, give me a pre-install risk pre-check list. Do not run commands for me; only explain what I should check, why, and what impact a failure would have.
```

### Host AI kickoff instruction

- Goal: Turn the project context into a host AI instruction for the start of a conversation.
- Expected output: A pre-work instruction with clear boundaries and clear evidence citations, suitable to copy to a host AI.

```text
Based on the AI Context Pack for neurostack, generate a pre-work instruction I can paste to my host AI. This instruction must obey not_runtime=true and must not claim the project has been installed, run, or produced real results.
```

## Role / Skill Index

- Indexed 54 role / Skill / project-doc entries.

- **NeuroStack - Claude Code Guide** (project_doc): NeuroStack is a neuroscience-grounded knowledge management system. CLI + MCP server + OpenAI-compatible API. Everything runs locally against a Markdown vault indexed in SQLite + FTS5. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CLAUDE.md`
- **🧠 neurostack - Build Your Personal Knowledge Vault** (project_doc): 🧠 neurostack - Build Your Personal Knowledge Vault Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `README.md`
- **NeuroStack** (project_doc): Build, maintain, and search your knowledge vault with AI. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `npm/README.md`
- **Vault — AI Agent Instructions** (project_doc): This is your knowledge base. AI agents should consult it before relying on general knowledge for any system-specific topic. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/AGENTS.md`
- **Contributing to NeuroStack** (project_doc): Thanks for your interest in contributing! Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CONTRIBUTING.md`
- **Neuroscience Appendix** (project_doc): NeuroStack's features are grounded in memory neuroscience. This appendix maps each feature to its scientific basis. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `docs/neuroscience-appendix.md`
- **NeuroStack Contributor License Agreement** (project_doc): NeuroStack Contributor License Agreement Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CLA.md`
- **Code of Conduct** (project_doc): We pledge to make participation in our community a harassment-free experience for everyone, regardless of age, body size, disability, ethnicity, gender identity, level of experience, nationality, personal appearance, race, religion, or sexual identity and orientation. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CODE_OF_CONDUCT.md`
- **Data Processing Agreement** (project_doc): Effective date: 25 March 2026 Last updated: 25 March 2026 Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `DPA.md`
- **Privacy Policy** (project_doc): Effective date: 25 March 2026 Last updated: 25 March 2026 Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `PRIVACY.md`
- **Security Policy** (project_doc): Version Supported --------- ----------- 0.1.x Yes Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `SECURITY.md`
- **memory-management** (project_doc): Create, update, merge, and manage persistent AI memories Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `src/neurostack/skills/memory-management.md`
- **session-lifecycle** (project_doc): Manage NeuroStack memory sessions for grouping memories Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `src/neurostack/skills/session-lifecycle.md`
- **vault-audit** (project_doc): Audit vault health - stale notes, missing summaries, prediction errors Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `src/neurostack/skills/vault-audit.md`
- **vault-search** (project_doc): Search the knowledge vault with the right retrieval strategy Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `src/neurostack/skills/vault-search.md`
- **Archive** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/archive/index.md`
- **Calendar** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/calendar/index.md`
- **Projects** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/home/projects/index.md`
- **Resources** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/home/resources/index.md`
- **Inbox** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/inbox/index.md`
- **Literature** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/literature/index.md`
- **Meta** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/meta/index.md`
- **Bias and Fairness** (project_doc): Bias in ML systems arises when model predictions systematically disadvantage specific groups. Fairness is not a single metric — it requires choosing which definition of fairness matches the deployment context. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/bias-and-fairness.md`
- **Data Versioning** (project_doc): Data versioning tracks the exact state of datasets, features, and model artefacts used in each experiment, making results reproducible and regressions traceable. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/data-versioning.md`
- **Experiment Tracking Tools** (project_doc): Experiment tracking systems log the inputs, parameters, metrics, and artefacts of every model training run, enabling comparison, reproducibility, and collaboration. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/experiment-tracking-tools.md`
- **Exploratory Data Analysis** (project_doc): EDA is the disciplined process of understanding a dataset's structure, quality, and distributions before modelling. Skipping EDA is the single most common source of avoidable model failures. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/exploratory-data-analysis.md`
- **Feature Engineering Patterns** (project_doc): Feature engineering is the process of transforming raw data into representations that improve model performance. The best features encode domain knowledge into a form the model can exploit. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/feature-engineering-patterns.md`
- **Research** (project_doc): - exploratory-data-analysis — Structured EDA process for new datasets - feature-engineering-patterns — Reusable feature transformations by data type - model-evaluation-metrics — Choosing the right metric for the problem - data-versioning — Tracking datasets and artefacts across experiments - experiment-tracking-tools — Comparing MLflow, W&B, DVC, and alternatives - bias-and-fairness — Measuring and mitigating bias i… Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/index.md`
- **Model Evaluation Metrics** (project_doc): Choosing the right evaluation metric is a modelling decision, not a technical one. The metric encodes what you care about — optimising the wrong metric produces a model that succeeds on paper and fails in production. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/research/model-evaluation-metrics.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/templates/analysis-note.md`
- **{{dataset}}** (project_doc): - Provider : {{source}} - Access : - License : Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/templates/dataset-note.md`
- **{{model}}** (project_doc): Parameter Value ----------- ------- Algorithm Learning rate Epochs / Iterations Regularisation Framework Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/templates/model-card.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/data-scientist/templates/pipeline-note.md`
- **API Design Principles** (project_doc): A well-designed API is easy to use correctly and hard to use incorrectly Bloch, 2006 . These principles apply to REST APIs, library interfaces, CLI tools, and internal module boundaries. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/api-design-principles.md`
- **Code Review Best Practices** (project_doc): Code review is one of the most effective defect-prevention techniques available, catching 60-90% of defects when done well Fagan, 1976; McConnell, 2004 . Its value extends beyond bug-finding to knowledge sharing, mentorship, and codebase consistency. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/code-review-best-practices.md`
- **Research** (project_doc): - twelve-factor-app — Methodology for building portable, resilient cloud-native services - technical-debt-management — Frameworks for identifying, quantifying, and paying down tech debt - code-review-best-practices — Evidence-based techniques for effective code review - testing-pyramid — Balancing test types for speed and confidence - api-design-principles — Designing APIs that are hard to misuse - refactoring-patte… Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/index.md`
- **Refactoring Patterns** (project_doc): Refactoring is the disciplined practice of restructuring existing code without changing its external behaviour Fowler, 1999 . The goal is to improve internal structure — readability, modularity, testability — while preserving correctness. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/refactoring-patterns.md`
- **Technical Debt Management** (project_doc): Technical debt is the implied cost of future rework caused by choosing a quick solution now instead of a better approach that would take longer Cunningham, 1992 . Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/technical-debt-management.md`
- **Twelve-Factor App** (project_doc): The twelve-factor methodology Wiggins, 2012 codifies best practices for building software-as-a-service applications that are portable, scalable, and suitable for continuous deployment. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/research/twelve-factor-app.md`
- **{{title}}** (project_doc): - Proposed - Accepted - Implemented - Superseded by Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/templates/architecture-decision.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/templates/code-review-note.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/templates/debugging-log.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/developer/templates/technical-spec.md`
- **Chaos Engineering** (project_doc): Chaos engineering is the discipline of experimenting on a system to build confidence in its ability to withstand turbulent conditions in production. It is not random destruction — it is disciplined, hypothesis-driven failure injection. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/chaos-engineering.md`
- **GitOps Workflow** (project_doc): GitOps is an operational framework where the entire system is described declaratively in git, and an automated agent ensures the live system matches the desired state in the repository. Git becomes the single source of truth for both application and infrastructure. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/gitops-workflow.md`
- **Incident Management Lifecycle** (project_doc): Incident management is the structured process of detecting, responding to, and learning from service disruptions. The goal is to restore service quickly and prevent recurrence through systemic improvements, not blame. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/incident-management-lifecycle.md`
- **Research** (project_doc): - sre-golden-signals — The four key metrics for monitoring any service - infrastructure-as-code-principles — Declarative, idempotent, version-controlled infrastructure - incident-management-lifecycle — From detection through blameless postmortem - gitops-workflow — Git as the single source of truth for deployments - observability-three-pillars — Logs, metrics, and traces as complementary signals - chaos-engineering… Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/index.md`
- **Infrastructure as Code Principles** (project_doc): Infrastructure as Code IaC is the practice of managing infrastructure through declarative definition files rather than manual processes, enabling version control, peer review, and reproducible environments. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/infrastructure-as-code-principles.md`
- **Observability: Three Pillars** (project_doc): Observability is the ability to understand a system's internal state from its external outputs. The three pillars — metrics, logs, and traces — are complementary signals that together provide the data needed to debug any production issue. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/observability-three-pillars.md`
- **SRE Golden Signals** (project_doc): The four golden signals from Google's SRE book are the minimum viable monitoring for any user-facing service. If you can only instrument four things, instrument these. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/research/sre-golden-signals.md`
- **{{title}}** (project_doc): - Impact : - Likelihood of failure : - Blast radius : Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/templates/change-request.md`
- **{{title}}** (project_doc): Time UTC Event ------------ ------- Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/templates/incident-report.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/templates/infrastructure-note.md`
- **{{title}}** (project_doc):  Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `vault-template/professions/devops/templates/runbook.md`

## Evidence Index

- Indexed 78 evidence entries.

- **NeuroStack - Claude Code Guide** (documentation): NeuroStack is a neuroscience-grounded knowledge management system. CLI + MCP server + OpenAI-compatible API. Everything runs locally against a Markdown vault indexed in SQLite + FTS5. Evidence: `CLAUDE.md`
- **🧠 neurostack - Build Your Personal Knowledge Vault** (documentation): 🧠 neurostack - Build Your Personal Knowledge Vault Evidence: `README.md`
- **NeuroStack** (documentation): Build, maintain, and search your knowledge vault with AI. Evidence: `npm/README.md`
- **Vault — AI Agent Instructions** (documentation): This is your knowledge base. AI agents should consult it before relying on general knowledge for any system-specific topic. Evidence: `vault-template/AGENTS.md`
- **Package** (package_manifest): { "name": "neurostack", "version": "0.10.1", "description": "Build, maintain, and search your knowledge vault with AI", "bin": { "neurostack": "bin/neurostack.js", "ns": "bin/neurostack.js" }, "scripts": { "postinstall": "node postinstall.js", "preuninstall": "node preuninstall.js" }, "keywords": "knowledge-management", "obsidian", "mcp", "ai", "rag", "sqlite", "fts5", "semantic-search", "knowledge-graph", "neuroscience", "cli" , "author": "Raphael Southall ", "license": "Apache-2.0", "repository": { "type": "git", "url": "https://github.com/raphasouthall/neurostack.git" }, "mcpName": "io.github.raphasouthall/neurostack", "homepage": "https://neurostack.sh", "bugs": { "url": "https://github… Evidence: `npm/package.json`
- **Contributing to NeuroStack** (documentation): Thanks for your interest in contributing! Evidence: `CONTRIBUTING.md`
- **License** (source_file): Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ Evidence: `LICENSE`
- **License** (source_file): Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ Evidence: `npm/LICENSE`
- **Neuroscience Appendix** (documentation): NeuroStack's features are grounded in memory neuroscience. This appendix maps each feature to its scientific basis. Evidence: `docs/neuroscience-appendix.md`
- **Server** (structured_config): { "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json", "name": "io.github.raphasouthall/neurostack", "description": "Token-efficient MCP memory for Markdown vaults. Tiered retrieval, hybrid search, GraphRAG.", "repository": { "url": "https://github.com/raphasouthall/neurostack", "source": "github" }, "version": "0.9.3", "packages": { "registryType": "npm", "identifier": "neurostack", "version": "0.9.3", "transport": { "type": "stdio" }, "environmentVariables": { "description": "Path to your Markdown vault", "isRequired": false, "format": "string", "isSecret": false, "name": "NEUROSTACK VAULT ROOT" } } } Evidence: `server.json`
- **Check PATH** (source_file): set -euo pipefail REPO="https://github.com/raphasouthall/neurostack.git" INSTALL DIR="${NEUROSTACK INSTALL DIR:-$HOME/.local/share/neurostack/repo}" CONFIG DIR="$HOME/.config/neurostack" MODE="${NEUROSTACK MODE:-lite}" info { echo " $ "; } warn { echo " ! $ " &2; } error { echo " X $ " &2; exit 1; } case "$ uname -s " in Linux info "Linux detected" ;; Darwin warn "macOS support is experimental" ;; error "Unsupported OS: $ uname -s . NeuroStack requires Linux." ;; esac PYTHON="" for cmd in python3.13 python3.12 python3.11 python3; do if command -v "$cmd" & /dev/null; then ver=$ "$cmd" -c "import sys; print f'{sys.version info.major}.{sys.version info.minor}' " major="${ver%%. }" minor="${ver… Evidence: `install.sh`
- **Core dependencies — lite mode FTS5 only, no ML** (source_file): project name = "neurostack" version = "0.10.1" description = "Build, maintain, and search your knowledge vault. CLI + MCP server with stale note detection, semantic search, and neuroscience-grounded memory." readme = "README.md" license = "Apache-2.0" requires-python = " =3.11" authors = { name = "Raphael Southall" }, keywords = "knowledge-management", "obsidian", "mcp", "neuroscience", "pkm", "zettelkasten", "rag", "local-first", "knowledge-graph", "semantic-search", "ai-memory", "second-brain" classifiers = "Development Status :: 3 - Alpha", "Environment :: Console", "Intended Audience :: Developers", "Intended Audience :: Science/Research", "Intended Audience :: Education", "License :: O… Evidence: `pyproject.toml`
- **Init** (source_file): version = "0.10.1" Evidence: `src/neurostack/__init__.py`
- **Api** (source_file): log = logging.getLogger "neurostack" ⋮---- MODELS = ⋮---- MODEL IDS = {m "id" for m in MODELS} ⋮---- class ChatMessage BaseModel ⋮---- role: str content: str ⋮---- class ChatCompletionRequest BaseModel ⋮---- model: str = "neurostack-ask" messages: list ChatMessage temperature: float = 0.3 max tokens: int None = None stream: bool = False top k: int = 8 workspace: str None = None ⋮---- class EmbeddingRequest BaseModel ⋮---- input: str list str model: str = "nomic-embed-text" encoding format: str = "float" ⋮---- def get api key - str ⋮---- def verify auth request: Request - None ⋮---- api key = get api key ⋮---- auth = request.headers.get "Authorization", "" ⋮---- token = auth len "Bearer " :… Evidence: `src/neurostack/api.py`
- **Split oversized chunks** (source_file): @dataclass class Chunk ⋮---- heading path: str content: str position: int ⋮---- @dataclass class ParsedNote ⋮---- path: str title: str frontmatter: dict = field default factory=dict content hash: str = "" chunks: list Chunk = field default factory=list wiki links: list str = field default factory=list ⋮---- FRONTMATTER RE = re.compile r"^---\s \n . ? \n---\s \n", re.DOTALL ⋮---- Split oversized chunks Evidence: `src/neurostack/chunker.py`
- **Try as path first** (source_file): def cmd index args ⋮---- db path = Path os.environ.get "NEUROSTACK DB PATH", DB PATH conn = get db db path notes = conn.execute "SELECT COUNT FROM notes" .fetchone 0 chunks = conn.execute "SELECT COUNT FROM chunks" .fetchone 0 edges = conn.execute "SELECT COUNT FROM graph edges" .fetchone 0 ⋮---- def get workspace args - str None ⋮---- ws = getattr args, "workspace", None ⋮---- ws = os.environ.get "NEUROSTACK WORKSPACE" ⋮---- def cmd search args ⋮---- results = hybrid search ⋮---- output = ⋮---- entry = { ⋮---- db path = Path os.environ.get "NEUROSTACK DB PATH", "" ⋮---- db path = Path DB PATH ⋮---- count = conn.execute "SELECT COUNT FROM notes" .fetchone 0 ⋮---- def cmd ask args ⋮---- resu… Evidence: `src/neurostack/cli.py`
- **Config** (source_file): @dataclass class CloudConfig ⋮---- cloud api url: str = "" cloud api key: str = "" ⋮---- def read toml - dict ⋮---- """Read existing config.toml or return empty dict.""" ⋮---- def write toml data: dict - None ⋮---- def save cloud config cloud api url: str, cloud api key: str - None ⋮---- data = read toml cloud = data.get "cloud", {} ⋮---- def clear cloud credentials - None ⋮---- def load cloud config - CloudConfig ⋮---- cfg = CloudConfig ⋮---- cloud data = data.get "cloud", {} ⋮---- env map = { ⋮---- val = os.environ.get env key Evidence: `src/neurostack/cloud/config.py`
- **Module-level singleton** (source_file): CONFIG PATH = Path.home / ".config" / "neurostack" / "config.toml" ⋮---- @dataclass class Config ⋮---- vault root: Path = field default factory=lambda: Path.home / "brain" db dir: Path = field default factory=lambda: Path.home / ".local" / "share" / "neurostack" embed url: str = "http://localhost:11435" embed model: str = "nomic-embed-text" embed dim: int = 768 llm url: str = "http://localhost:11434" ⋮---- llm model: str = "phi3.5" llm api key: str = "" embed api key: str = "" session dir: Path = field default factory=lambda: Path.home / ".claude" / "projects" api host: str = "127.0.0.1" api port: int = 8000 api key: str = "" cooccurrence boost weight: float = 0.1 ⋮---- @property def db pat… Evidence: `src/neurostack/config.py`
- **Cooccurrence** (source_file): log = logging.getLogger "neurostack" ⋮---- MAX COOCCURRENCE WEIGHT = 100.0 ⋮---- canonical: set tuple str, str = set ⋮---- now = datetime.now timezone.utc .isoformat updates: list tuple str, str, float, str = ⋮---- row = conn.execute ⋮---- new weight = min row "weight" 1.1, MAX COOCCURRENCE WEIGHT ⋮---- new weight = 1.0 ⋮---- def persist cooccurrence conn: sqlite3.Connection - int ⋮---- rows = conn.execute ⋮---- note entities: dict str, set str = defaultdict set ⋮---- pair weights: dict tuple str, str , float = defaultdict float ⋮---- entity list = sorted entities ⋮---- def upsert cooccurrence for note conn: sqlite3.Connection, note path: str - int ⋮---- """Incrementally update co-occurrenc… Evidence: `src/neurostack/cooccurrence.py`
- **Embedder** (source_file): HAS NUMPY = True ⋮---- HAS NUMPY = False ⋮---- cfg = get config DEFAULT EMBED URL = cfg.embed url EMBED MODEL = cfg.embed model EMBED DIM = cfg.embed dim EMBED HEADERS = auth headers cfg.embed api key ⋮---- payload = {"model": model, "input": text} ⋮---- resp = httpx.post ⋮---- data = resp.json ⋮---- all embeddings = ⋮---- batch = texts i : i + batch size payload = {"model": model, "input": batch} ⋮---- def embedding to blob vec: "np.ndarray" - bytes ⋮---- def blob to embedding blob: bytes - "np.ndarray" ⋮---- header = f"Note: {title}" ⋮---- fm = json.loads frontmatter json if frontmatter json else {} ⋮---- tags = fm.get "tags" or fm.get "tag" ⋮---- parts = header ⋮---- def cosine similarit… Evidence: `src/neurostack/embedder.py`
- **Use the resolved path from DB** (source_file): @dataclass class GraphNode ⋮---- path: str title: str summary: str pagerank: float in degree: int out degree: int ⋮---- @dataclass class GraphResult ⋮---- center: GraphNode neighbors: list GraphNode ⋮---- def resolve wiki link link target: str, all paths: list str - str None ⋮---- target lower = link target.lower .strip ⋮---- stem = Path p .stem.lower ⋮---- def build graph conn: sqlite3.Connection, vault root: Path ⋮---- all paths = r "path" for r in conn.execute "SELECT path FROM notes" .fetchall ⋮---- full path = vault root / note path ⋮---- text = full path.read text encoding="utf-8", errors="replace" links = extract wiki links text ⋮---- target = resolve wiki link link, all paths ⋮----… Evidence: `src/neurostack/graph.py`
- **Related** (source_file): HAS NUMPY = True ⋮---- HAS NUMPY = False ⋮---- log = logging.getLogger "neurostack" ⋮---- conn = get db DB PATH workspace = normalize workspace workspace ⋮---- source rows = conn.execute ⋮---- source embeddings = blob to embedding r "embedding" for r in source rows source mean = np.mean np.stack source embeddings , axis=0 ⋮---- all rows = conn.execute ⋮---- note embeddings: dict str, list np.ndarray = {} ⋮---- np = r "note path" ⋮---- scored: list tuple str, float = ⋮---- note mean = np.mean np.stack embs , axis=0 sim = cosine similarity source mean, note mean ⋮---- top = scored :top k ⋮---- results = ⋮---- note row = conn.execute title = note row "title" if note row else np ⋮---- summary r… Evidence: `src/neurostack/related.py`
- **Server** (source_file): mcp = create mcp server Evidence: `src/neurostack/server.py`
- **Init** (source_file): registered = False ⋮---- def ensure registered - ToolRegistry ⋮---- registered = True ⋮---- all = Evidence: `src/neurostack/tools/__init__.py`
- **Mcp Adapter** (source_file): log = logging.getLogger "neurostack.tools.mcp adapter" ⋮---- def create mcp server name: str = "neurostack", fastmcp kwargs - FastMCP ⋮---- mcp = FastMCP name, fastmcp kwargs registry = ensure registered ⋮---- @functools.wraps tool def.fn def wrapper td=tool def, kwargs Evidence: `src/neurostack/tools/mcp_adapter.py`
- **Memory Tools** (source_file): def embed url ⋮---- conn = get db DB PATH memory = save memory ⋮---- result = { ⋮---- @registry.tool tags= "memory", "write" def vault forget memory id: int - dict ⋮---- deleted = forget memory conn, memory id ⋮---- memory = update memory ⋮---- changed = ⋮---- memory = merge memories ⋮---- memories = search memories ⋮---- output = ⋮---- entry = { ⋮---- vault root = get config .vault root Evidence: `src/neurostack/tools/memory_tools.py`
- **Openai Adapter** (source_file): log = logging.getLogger "neurostack.tools.openai adapter" ⋮---- TYPE MAP: dict type str, str = { ⋮---- def param to json schema param: ToolParam - dict str, Any ⋮---- ptype = param.type schema: dict str, Any = {} ⋮---- origin = getattr ptype, " origin ", None ⋮---- args = getattr ptype, " args ", ⋮---- item type = TYPE MAP.get args 0 , "string" ⋮---- def tool to openai function tool: ToolDef - dict str, Any ⋮---- properties: dict str, Any = {} required: list str = ⋮---- parameters: dict str, Any = { ⋮---- doc = inspect.getdoc tool.fn or tool.description ⋮---- doc = doc :1021 + "..." ⋮---- def get openai tools tag: str None = None - list dict str, Any ⋮---- registry = ensure registered ⋮----… Evidence: `src/neurostack/tools/openai_adapter.py`
- **First paragraph up to blank line** (source_file): log = logging.getLogger "neurostack.tools" ⋮---- @dataclass frozen=True class ToolParam ⋮---- name: str type: type str default: Any = inspect.Parameter.empty description: str = "" ⋮---- @property def required self - bool ⋮---- @dataclass frozen=True class ToolDef ⋮---- """Complete definition of a registered tool.""" ⋮---- description: str fn: Callable ..., dict params: list ToolParam tags: tuple str, ... = ⋮---- def call self, kwargs: Any - dict ⋮---- """Invoke the tool function with the given kwargs.""" ⋮---- def extract params fn: Callable - list ToolParam ⋮---- """Extract parameter metadata from a function signature + type hints.""" sig = inspect.signature fn ⋮---- hints = get type hints… Evidence: `src/neurostack/tools/registry.py`
- **Rest Adapter** (source_file): log = logging.getLogger "neurostack.tools.rest adapter" ⋮---- def create tools router prefix: str = "/v1/tools" - APIRouter ⋮---- router = APIRouter prefix=prefix, tags= "tools" registry = ensure registered ⋮---- @router.get "", summary="List all available tools" async def list tools - JSONResponse ⋮---- tools = ⋮---- properties: dict str, Any = {} required: list str = ⋮---- schema: dict str, Any = { ⋮---- async def invoke tool tool name: str, request: Request - JSONResponse ⋮---- tool = registry.get tool name ⋮---- body = await request.json ⋮---- body = {} ⋮---- result = tool.call body Evidence: `src/neurostack/tools/rest_adapter.py`
- **Validate structure** (source_file): log = logging.getLogger "neurostack" ⋮---- cfg = get config DEFAULT SUMMARIZE URL = cfg.llm url TRIPLE MODEL = cfg.llm model LLM HEADERS = auth headers cfg.llm api key ⋮---- TRIPLE PROMPT = """Extract knowledge graph triples from this note. \ ⋮---- content = content :4000 + "\n ... truncated " ⋮---- prompt = TRIPLE PROMPT.format title=title, content=content ⋮---- resp = httpx.post ⋮---- data = resp.json raw = data "choices" 0 "message" "content" .strip ⋮---- raw = re.sub r"^ $", "", raw .strip ⋮---- triples = json.loads raw ⋮---- Validate structure valid = ⋮---- s = str t "s" .strip p = str t "p" .strip o = str t "o" .strip ⋮---- def triple to text t: dict - str Evidence: `src/neurostack/triples.py`
- **NeuroStack Contributor License Agreement** (documentation): NeuroStack Contributor License Agreement Evidence: `CLA.md`
- **Code of Conduct** (documentation): We pledge to make participation in our community a harassment-free experience for everyone, regardless of age, body size, disability, ethnicity, gender identity, level of experience, nationality, personal appearance, race, religion, or sexual identity and orientation. Evidence: `CODE_OF_CONDUCT.md`
- **Data Processing Agreement** (documentation): Effective date: 25 March 2026 Last updated: 25 March 2026 Evidence: `DPA.md`
- **Privacy Policy** (documentation): Effective date: 25 March 2026 Last updated: 25 March 2026 Evidence: `PRIVACY.md`
- **Security Policy** (documentation): Version Supported --------- ----------- 0.1.x Yes Evidence: `SECURITY.md`
- **Memory Management** (documentation): Merge duplicate memories Keeps the longer content, unions tags, picks the more specific entity type. Evidence: `src/neurostack/skills/memory-management.md`
- **Session Lifecycle** (documentation): Sessions group memories created during a conversation for later review. Evidence: `src/neurostack/skills/session-lifecycle.md`
- **Vault Audit** (documentation): Quick health check Shows note count, chunk count, memory count, embedding coverage. Evidence: `src/neurostack/skills/vault-audit.md`
- **Vault Search Guide** (documentation): NeuroStack has multiple retrieval tools. Pick the right one: Evidence: `src/neurostack/skills/vault-search.md`
- **Archive** (documentation): Archive Evidence: `vault-template/archive/index.md`
- **Calendar** (documentation): Calendar Evidence: `vault-template/calendar/index.md`
- **Projects** (documentation): Projects Evidence: `vault-template/home/projects/index.md`
- **Resources** (documentation): Resources Evidence: `vault-template/home/resources/index.md`
- **Inbox** (documentation): Inbox Evidence: `vault-template/inbox/index.md`
- **Literature** (documentation): Literature Evidence: `vault-template/literature/index.md`
- **Meta** (documentation): Meta Evidence: `vault-template/meta/index.md`
- **Bias and Fairness** (documentation): Bias in ML systems arises when model predictions systematically disadvantage specific groups. Fairness is not a single metric — it requires choosing which definition of fairness matches the deployment context. Evidence: `vault-template/professions/data-scientist/research/bias-and-fairness.md`
- **Data Versioning** (documentation): Data versioning tracks the exact state of datasets, features, and model artefacts used in each experiment, making results reproducible and regressions traceable. Evidence: `vault-template/professions/data-scientist/research/data-versioning.md`
- **Experiment Tracking Tools** (documentation): Experiment tracking systems log the inputs, parameters, metrics, and artefacts of every model training run, enabling comparison, reproducibility, and collaboration. Evidence: `vault-template/professions/data-scientist/research/experiment-tracking-tools.md`
- **Exploratory Data Analysis** (documentation): EDA is the disciplined process of understanding a dataset's structure, quality, and distributions before modelling. Skipping EDA is the single most common source of avoidable model failures. Evidence: `vault-template/professions/data-scientist/research/exploratory-data-analysis.md`
- **Feature Engineering Patterns** (documentation): Feature engineering is the process of transforming raw data into representations that improve model performance. The best features encode domain knowledge into a form the model can exploit. Evidence: `vault-template/professions/data-scientist/research/feature-engineering-patterns.md`
- **Research** (documentation): - exploratory-data-analysis — Structured EDA process for new datasets - feature-engineering-patterns — Reusable feature transformations by data type - model-evaluation-metrics — Choosing the right metric for the problem - data-versioning — Tracking datasets and artefacts across experiments - experiment-tracking-tools — Comparing MLflow, W&B, DVC, and alternatives - bias-and-fairness — Measuring and mitigating bias in ML systems Evidence: `vault-template/professions/data-scientist/research/index.md`
- **Model Evaluation Metrics** (documentation): Choosing the right evaluation metric is a modelling decision, not a technical one. The metric encodes what you care about — optimising the wrong metric produces a model that succeeds on paper and fails in production. Evidence: `vault-template/professions/data-scientist/research/model-evaluation-metrics.md`
- **{{title}}** (documentation): --- date: {{date}} tags: analysis type: project status: active actionable: true compositional: true --- {{title}} Objective Data Source Methodology Key Findings 1. 2. 3. Visualisations Limitations Recommendations - Related Notes Evidence: `vault-template/professions/data-scientist/templates/analysis-note.md`
- **{{dataset}}** (documentation): - Provider : {{source}} - Access : - License : Evidence: `vault-template/professions/data-scientist/templates/dataset-note.md`
- **{{model}}** (documentation): Parameter Value ----------- ------- Algorithm Learning rate Epochs / Iterations Regularisation Framework Evidence: `vault-template/professions/data-scientist/templates/model-card.md`
- **{{title}}** (documentation): --- date: {{date}} tags: pipeline type: project status: active actionable: true compositional: false --- {{title}} Purpose Architecture transform - sink -- Inputs Transformations 1. 2. 3. Outputs Schedule & Triggers - Frequency : - Trigger : - SLA : Monitoring & Alerts Dependencies Related Notes Evidence: `vault-template/professions/data-scientist/templates/pipeline-note.md`
- **API Design Principles** (documentation): A well-designed API is easy to use correctly and hard to use incorrectly Bloch, 2006 . These principles apply to REST APIs, library interfaces, CLI tools, and internal module boundaries. Evidence: `vault-template/professions/developer/research/api-design-principles.md`
- **Code Review Best Practices** (documentation): Code review is one of the most effective defect-prevention techniques available, catching 60-90% of defects when done well Fagan, 1976; McConnell, 2004 . Its value extends beyond bug-finding to knowledge sharing, mentorship, and codebase consistency. Evidence: `vault-template/professions/developer/research/code-review-best-practices.md`
- **Research** (documentation): - twelve-factor-app — Methodology for building portable, resilient cloud-native services - technical-debt-management — Frameworks for identifying, quantifying, and paying down tech debt - code-review-best-practices — Evidence-based techniques for effective code review - testing-pyramid — Balancing test types for speed and confidence - api-design-principles — Designing APIs that are hard to misuse - refactoring-patterns — Safe, incremental strategies for improving existing code Evidence: `vault-template/professions/developer/research/index.md`
- The remaining 18 evidence entries are in `AI_CONTEXT_PACK.json` or `EVIDENCE_INDEX.json`.

## Rules the Host AI Must Follow

- **Treat this asset as pre-work context, not a runtime environment.**: The AI Context Pack contains only an evidence-backed understanding of the project, not the project's executable state. Evidence: `CLAUDE.md`, `README.md`, `npm/README.md`
- **When answering the user, distinguish what can be previewed from what can only be verified after install.**: The consumer value of the pre-install experience comes from reducing bad installs and misjudgments, not from pretending to be a real run. Evidence: `CLAUDE.md`, `README.md`, `npm/README.md`

## Questions the User Should Answer First

- Which host AI or local environment do you plan to use it in?
- Do you just want to experience the workflow first, or are you ready to actually install?
- What matters most to you: install cost, output quality, or conflicts with your existing rules?

## Acceptance Checks

- Every capability claim can be traced back to a file path in evidence_refs.
- AI_CONTEXT_PACK.md does not package previews as a real run.
- The user can understand who it fits, what it can do, how to start, and the risk boundaries within 3 minutes.

---

## Doramagic Context Augmentation

The following sections strengthen the repository context for a host AI. Human Manual data is a reading route, and pitfall notes become operating constraints.

## Human Manual Outline

Usage rule: this is only a reading route and salience signal, not factual authority. Concrete claims must still return to repo evidence or Claim Graph.

Host AI hard rules:
- Do not treat page titles, section order, summaries, or importance values as factual project evidence.
- When explaining the Human Manual outline, state that it is only a reading route or salience signal.
- Capability, installation, compatibility, runtime state, and risk claims must cite repo evidence, source paths, or Claim Graph.

- **Introduction to neurostack**: importance `high`
  - source_paths: README.md, CLAUDE.md, server.json, pyproject.toml, install.sh
- **System Architecture and Module Layout**: importance `high`
  - source_paths: src/neurostack/__init__.py, src/neurostack/__main__.py, src/neurostack/cli.py, src/neurostack/api.py, src/neurostack/server.py
- **Memory, Search, and Knowledge Graph**: importance `high`
  - source_paths: src/neurostack/chunker.py, src/neurostack/embedder.py, src/neurostack/graph.py, src/neurostack/cooccurrence.py, src/neurostack/triples.py
- **AI Adapters, Cloud Sync, Skills, and Templates**: importance `high`
  - source_paths: src/neurostack/tools/__init__.py, src/neurostack/tools/registry.py, src/neurostack/tools/openai_adapter.py, src/neurostack/tools/mcp_adapter.py, src/neurostack/tools/rest_adapter.py

## Repo Inspection Evidence

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `0a6700b1a5da5088ae6f3e79c782778311e2445c`
- inspected_files: `README.md`, `pyproject.toml`, `docs/neuroscience-appendix.md`, `src/neurostack/__init__.py`, `src/neurostack/__main__.py`, `src/neurostack/api.py`, `src/neurostack/ask.py`, `src/neurostack/attractor.py`, `src/neurostack/brief.py`, `src/neurostack/bundle.py`, `src/neurostack/capture.py`, `src/neurostack/chunker.py`, `src/neurostack/cli.py`, `src/neurostack/cloud/__init__.py`, `src/neurostack/cloud/client.py`, `src/neurostack/cloud/config.py`, `src/neurostack/cloud/manifest.py`, `src/neurostack/cloud/sync.py`, `src/neurostack/community.py`, `src/neurostack/community_search.py`

Host AI hard rules:
- Without repo_clone_verified=true, do not claim that the source code has been read.
- Without repo_inspection_verified=true, do not write README, docs, or package-file conclusions as facts.
- Without quick_start_verified=true, do not claim that the Quick Start path has run successfully.

## Doramagic Pitfall Constraints

These rules come from Doramagic discovery, validation, or compilation findings. The host AI must treat them as operating constraints, not background notes.

### Constraint 1: Capability evidence risk requires verification

- Trigger: README/documentation is current enough for a first validation pass.
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: capability.assumptions | https://github.com/Abhinav1234abhinav/neurostack
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 2: Security or permission risk requires verification

- Trigger: no_demo
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: downstream_validation.risk_items | https://github.com/Abhinav1234abhinav/neurostack
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 3: Security or permission risk requires verification

- Trigger: no_demo
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: risks.scoring_risks | https://github.com/Abhinav1234abhinav/neurostack
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 4: Maintenance risk requires verification

- Trigger: issue_or_pr_quality=unknown。
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: evidence.maintainer_signals | https://github.com/Abhinav1234abhinav/neurostack
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 5: Maintenance risk requires verification

- Trigger: release_recency=unknown。
- Host AI rule: Reproduce the official install and quickstart path in an isolated environment.
- Why it matters: May increase setup, validation, or first-run risk for the user.
- Evidence: evidence.maintainer_signals | https://github.com/Abhinav1234abhinav/neurostack
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.