# bastra-recall - 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 bastra-recall. 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

- **Developers already using host AIs such as Claude/Codex/Cursor/Gemini**: The README or plugin config mentions multiple host AIs. Evidence: `README.md` Claim: `clm_0003` supported 0.86
- **Users who want to bring professional workflows into a host AI**: The repo contains Skill documents. Evidence: `packages/skill/SKILL.md` Claim: `clm_0004` supported 0.86

## What It Can Do

- **AI Skill / Agent Instruction Asset Library** (Previewable before install): The project contains Skill or Agent instruction files that a host AI can read, useful for bringing professional workflows into hosts like Claude, Codex, or Cursor. Evidence: `packages/skill/SKILL.md` Claim: `clm_0001` supported 0.86
- **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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.md` Claim: `clm_0002` supported 0.86

## How to Start

- `git clone https://github.com/n0mad-ai/bastra-recall.git` Evidence: `README.md` Claim: `clm_0005` supported 0.86
- `npx bastra-recall install           # zero-install: guided setup with selection lists` Evidence: `README.md` Claim: `clm_0006` supported 0.86
- `npx bastra-recall install all       # direct: all clients (script-friendly, add --yes for CI)` Evidence: `README.md` Claim: `clm_0007` supported 0.86
- `npm install -g bastra-recall` Evidence: `README.md` Claim: `clm_0008` supported 0.86
- `npx bastra-recall install           # ohne Installation: geführtes Setup mit Auswahllisten` Evidence: `README.md` Claim: `clm_0009` supported 0.86
- `npx bastra-recall install all       # direkt: alle Clients (skript-tauglich, --yes für CI)` Evidence: `README.md` Claim: `clm_0010` supported 0.86
- `npx bastra-recall install` Evidence: `packages/bastra-recall/README.md` Claim: `clm_0006` supported 0.86, `clm_0007` supported 0.86, `clm_0009` supported 0.86, `clm_0010` supported 0.86 et al.
- `curl -sS http://127.0.0.1:6723/health` Evidence: `packages/daemon/README.md` Claim: `clm_0012` 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**: Tool permission boundaries cannot be trusted before install.
- **Continuing will touch**: Command execution, Host AI configuration, Local environment or project files

### What You Can Trust Now

- **Target-audience signal: Developers already using host AIs such as Claude/Codex/Cursor/Gemini** (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_0003` supported 0.86
- **Target-audience signal: Users who want to bring professional workflows into a host AI** (supported): Backed by a supported claim or project evidence, but that still is not the same as real install results. Evidence: `packages/skill/SKILL.md` Claim: `clm_0004` supported 0.86
- **Capability exists: AI Skill / Agent Instruction Asset Library** (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: `packages/skill/SKILL.md` Claim: `clm_0001` 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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.md` Claim: `clm_0002` 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: `README.md` Claim: `clm_0005` supported 0.86

### What You Cannot Trust Yet

- **Tool permission boundaries cannot be trusted before install.** (unverified): MCP/tool projects usually touch files, the network, the browser, or external APIs, so permissions and logs must be checked for real.
- **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: `packages/skill/SKILL.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.
- **Do the install commands require network access, permissions, or global writes?** (unverified): This affects install risk in both enterprise and personal environments. Evidence: `README.md`

### What Continuing Will Touch

- **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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.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: `packages/skill/SKILL.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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.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: `README.md`, `docs/architecture.md`, `packages/daemon/README.md`
- **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 a pre-install interactive trial to judge whether the way of working fits; it needs no authorization or environment change. (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.
- **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_0013` 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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.md` Claim: `clm_0014` 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

- **AI Skill / Agent Instruction Asset Library**: Use role_skill_index / evidence_index to help the user pick a usable role, Skill, or workflow first. Boundary: Can be experienced via a pre-install Prompt. Evidence: `packages/skill/SKILL.md` Claim: `clm_0001` supported 0.86
- **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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/README.md` Claim: `clm_0002` supported 0.86

### Context Scale

- Total files: 275
- Important-file coverage: 40/275
- Evidence index entries: 66
- Role / Skill entries: 1

### 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 bastra-recall, 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 bastra-recall 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 bastra-recall, 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 1 role / Skill / project-doc entries.

- **bastra-recall** (skill): Persistent external brain for Claude — documents PDFs, contracts, scans with OCR , personal facts appointments, decisions, items, amounts , AND code lessons / preferences / project topology. USE PROACTIVELY in three modes. 1 RECALL — whenever the user asks about anything from their past, vault, projects, or personal life, INCLUDING direct retrieval phrasings like "find...", "where is...", "when was...", "how much wa… Activation hint: When the user's task is highly relevant to the workflow described by “bastra-recall”, use it for a pre-install experience first, then decide whether to install. Evidence: `packages/skill/SKILL.md`

## Evidence Index

- Indexed 66 evidence entries.

- **Bastra.Recall** (documentation): A persistent teammate memory for any AI assistant — across every surface. Ein persistentes Teammate-Gedächtnis für jeden AI-Assistenten — über jede Oberfläche hinweg. Evidence: `README.md`
- **bastra-recall** (documentation): Persistent recall memory for AI coding agents — Claude Code, Claude Desktop, and Cursor. This package is a thin launcher that installs and wires up bastra-recall across every MCP-capable client with one command. Evidence: `packages/bastra-recall/README.md`
- **@bastra-recall/daemon** (documentation): The MCP server + HTTP gateway behind bastra-recall. Watches a markdown vault, indexes it with BM25 + optional embeddings, and exposes a stable tool surface over both stdio MCP and HTTP REST. Evidence: `packages/daemon/README.md`
- **@bastra-recall/eval — recall when marginal-lift ablation** (documentation): @bastra-recall/eval — recall when marginal-lift ablation Evidence: `packages/eval/README.md`
- **Out-of-pool predictor + stratified transfer-gate** (documentation): Out-of-pool predictor + stratified transfer-gate Evidence: `packages/eval/oop-gate/README.md`
- **@bastra-recall/statusline** (documentation): Statusline for Claude Code, shipped with bastra-recall https://github.com/n0mad-ai/bastra-recall . Evidence: `packages/statusline/README.md`
- **Package** (package_manifest): { "name": "bastra-recall-monorepo", "version": "0.7.0-beta.5", "description": "Workspace root for bastra-recall: OSS daemon, MCP server, and Claude Code skill — the open foundation under the Bastra Mac app.", "private": true, "license": "MIT", "engines": { "node": " =22" }, "workspaces": "packages/core", "packages/daemon", "packages/statusline", "packages/bastra-recall", "packages/eval" , "scripts": { "build": "npm run build --workspaces --if-present", "check:types": "npm run check:types --workspaces --if-present", "test": "node --import tsx --test packages/core/ tests / .test.ts packages/daemon/ tests / .test.ts packages/eval/ tests / .test.ts", "eval:lift": "npm run lift --workspace=@bast… Evidence: `package.json`
- **Package** (package_manifest): { "name": "bastra-recall", "version": "0.7.0-beta.5", "description": "One-command installer for bastra-recall — persistent recall memory for Claude Code, Claude Desktop, and Cursor. Thin launcher for the bastra CLI.", "type": "module", "license": "MIT", "repository": { "type": "git", "url": "git+https://github.com/n0mad-ai/bastra-recall.git", "directory": "packages/bastra-recall" }, "publishConfig": { "access": "public" }, "engines": { "node": " =22" }, "bin": { "bastra": "./bin/bastra.mjs" }, "files": "bin", "README.md" , "dependencies": { "@bastra-recall/daemon": "0.7.0-beta.5" }, "keywords": "bastra", "recall", "memory", "mcp", "claude", "claude-code", "cursor" , "homepage": "https://bas… Evidence: `packages/bastra-recall/package.json`
- **Package** (package_manifest): { "name": "@bastra-recall/core", "version": "0.7.0-beta.5", "description": "Reusable core library: vault loader, BM25 index, save memory. Used by the daemon and the Mac-app.", "type": "module", "license": "MIT", "repository": { "type": "git", "url": "git+https://github.com/n0mad-ai/bastra-recall.git", "directory": "packages/core" }, "publishConfig": { "access": "public" }, "engines": { "node": " =22" }, "main": "./dist/index.js", "types": "./dist/index.d.ts", "exports": { ".": { "types": "./dist/index.d.ts", "import": "./dist/index.js" }, "./scrub": { "types": "./dist/scrub.d.ts", "import": "./dist/scrub.js" } }, "files": "dist", "src" , "scripts": { "build": "tsc", "check:types": "tsc --no… Evidence: `packages/core/package.json`
- **Package** (package_manifest): { "name": "@bastra-recall/daemon", "version": "0.7.0-beta.5", "description": "MCP server exposing recall + load memory over a markdown vault.", "type": "module", "license": "MIT", "repository": { "type": "git", "url": "git+https://github.com/n0mad-ai/bastra-recall.git", "directory": "packages/daemon" }, "publishConfig": { "access": "public" }, "engines": { "node": " =22" }, "bin": { "bastra": "./dist/cli.js", "bastra-recall": "./dist/index.js", "bastra-recall-mcp": "./dist/mcp-forwarder.js", "bastra-recall-bridge": "./dist/bridge.js", "bastra-recall-hook": "./dist/hook.js", "bastra-recall-session-hook": "./dist/session-hook.js", "bastra-recall-prompt-hook": "./dist/prompt-hook.js", "bastra-… Evidence: `packages/daemon/package.json`
- **Package** (package_manifest): { "name": "@bastra-recall/eval", "version": "0.7.0-beta.5", "description": "Offline ablation harness: marginal lift of the recall when field under paraphrased held-out queries control: stripped vs treatment: x5 .", "private": true, "type": "module", "license": "MIT", "engines": { "node": " =22" }, "scripts": { "lift": "tsx src/marginal-lift.ts", "personas": "tsx src/persona-lift.ts", "bridge-transfer": "tsx src/bridge-transfer.ts", "doc2query-lift": "tsx src/doc2query-lift.ts", "check:types": "tsc --noEmit" }, "dependencies": { "@bastra-recall/core": "0.7.0-beta.5", "minisearch": "^7.1.1" }, "devDependencies": { "@types/node": "^22.20.0", "tsx": "^4.19.2", "typescript": "^6.0.3" }, "homepag… Evidence: `packages/eval/package.json`
- **Package** (package_manifest): { "name": "@bastra-recall/statusline", "version": "0.7.0-beta.5", "description": "bastra-recall statusline for Claude Code. Statusline engine: owloops/claude-powerline MIT , with the bastra-status segment built in.", "type": "module", "license": "MIT", "repository": { "type": "git", "url": "git+https://github.com/n0mad-ai/bastra-recall.git", "directory": "packages/statusline" }, "publishConfig": { "access": "public" }, "engines": { "node": " =22" }, "main": "./dist/index.mjs", "bin": { "bastra-statusline": "./bin/claude-powerline" }, "files": "dist", "bin", "src", "README.md", "LICENSE" , "scripts": { "build": "tsdown", "check:types": "tsc --noEmit" }, "devDependencies": { "@types/node": "^… Evidence: `packages/statusline/package.json`
- **bastra-recall — autonomous teammate memory** (skill_instruction): bastra-recall — autonomous teammate memory Evidence: `packages/skill/SKILL.md`
- **Architecture** (documentation): Bastra.Recall is a local-first memory layer for AI assistants. It gives Claude Code, Claude Desktop, Cursor, ChatGPT Actions, and other MCP/HTTP clients one shared vault of durable lessons, preferences, project facts, decisions, workflows, bookmarks, and document sidecars. Evidence: `docs/architecture.md`
- **Bastra Commons — shared recall 119 / 120** (documentation): Bastra Commons — shared recall 119 / 120 Evidence: `docs/commons.md`
- **Eval geometry** (documentation): A pre-run discipline for the eval threads 89/ 102, 120/ 121, 130, 129, 134, 135 . Across those threads the same bug recurred — measuring the wrong region of the data — and it was almost always caught after a run. This note makes the boundary a pre-run object so a regime slip reads as a visible category error instead of a number that quietly goes wrong. Evidence: `docs/eval-geometry.md`
- **Claude Code hooks for bastra-recall** (documentation): Claude Code hooks for bastra-recall Evidence: `docs/hooks.md`
- **Memory Schema** (documentation): This schema defines one stored memory or document sidecar. Files are plain markdown with YAML frontmatter so they remain editable in Obsidian and by hand. Evidence: `docs/memory-schema.md`
- **Survival — the by-id substrate invariant** (documentation): Survival — the by-id substrate invariant Evidence: `docs/survival.md`
- **Self-learning taxonomy 64** (documentation): bastra-recall's category system has two kinds of axes: Evidence: `docs/taxonomy.md`
- **Triggers — when to save, when to recall** (documentation): Triggers — when to save, when to recall Evidence: `docs/triggers.md`
- **Claude Code: MCP notifications/progress arrive but are never rendered in the TUI** (documentation): Claude Code: MCP notifications/progress arrive but are never rendered in the TUI Evidence: `docs/issues/cc-mcp-progress-rendering.md`
- **Contributing to Bastra Recall** (documentation): Thanks for considering a contribution. Bastra is solo-maintained, so the workflow is light — but a few conventions keep things smooth. Evidence: `CONTRIBUTING.md`
- **License** (source_file): Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files the "Software" , to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: Evidence: `LICENSE`
- **License** (source_file): Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files the "Software" , to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: Evidence: `packages/eval/LICENSE`
- **License** (source_file): Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files the "Software" , to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: Evidence: `packages/statusline/LICENSE`
- **Changelog** (documentation): All notable changes to bastra-recall are documented here. The format is based on Keep a Changelog https://keepachangelog.com/ , and this project adheres to Semantic Versioning https://semver.org/ . Evidence: `CHANGELOG.md`
- **Bastra.Recall Roadmap** (documentation): Bastra.Recall is a local-first memory layer for AI assistants. The user should not have to re-state durable preferences, project facts, decisions, workflows, or lessons across Claude Code, Claude Desktop, Cursor, ChatGPT Actions, and other MCP/HTTP clients. Evidence: `PLAN.md`
- **!/usr/bin/env bash** (source_file): !/usr/bin/env bash Install Bastra.command — double-click in Finder to set up bastra-recall. Idempotent: safe to run multiple times. Installs Homebrew if missing, adds the bastra tap, installs bastra-recall, then hands over to the guided setup bastra install : selection lists for the memory vault, your AI clients, and semantic recall — no flags, no typing paths. After this script finishes, restart the AI client s you use — the memory tool will be live. Evidence: `distribution/Install Bastra.command`
- **Openapi** (source_file): openapi: 3.0.3 info: title: Bastra.Recall REST API version: 0.6.0-beta.1 description: Loopback REST API exposed by the bastra-recall daemon. For hosted clients such as ChatGPT Actions, expose 127.0.0.1:6723 through a tunnel and set BASTRA API TOKEN . servers: - url: http://127.0.0.1:6723/api/v1 security: - bearerAuth: paths: /recall: post: operationId: recall summary: Search memories requestBody: required: true content: application/json: schema: $ref: " /components/schemas/RecallRequest" responses: "200": description: Recall candidates content: application/json: schema: $ref: " /components/schemas/RecallResponse" /load memory: post: operationId: loadMemory summary: Load one memory by id req… Evidence: `docs/openapi.yaml`
- **Bastra Recall** (source_file): class BastraRecall "bastra" bin.install symlink libexec/"packages/daemon/dist/index.js" = "bastra-recall" bin.install symlink libexec/"packages/daemon/dist/mcp-forwarder.js" = "bastra-recall-mcp" bin.install symlink libexec/"packages/daemon/dist/bridge.js" = "bastra-recall-bridge" bin.install symlink libexec/"packages/daemon/dist/hook.js" = "bastra-recall-hook" bin.install symlink libexec/"packages/daemon/dist/session-hook.js" = "bastra-recall-session-hook" bin.install symlink libexec/"packages/daemon/dist/prompt-hook.js" = "bastra-recall-prompt-hook" bin.install symlink libexec/"packages/daemon/dist/todo-hook.js" = "bastra-recall-todo-hook" bin.install symlink libexec/"packages/daemon/dist… Evidence: `distribution/homebrew/bastra-recall.rb`
- **Bastra** (source_file): // Thin launcher for the unscoped bastra-recall package. // // Re-exports the daemon's bastra CLI so npx bastra-recall install all and // npm i -g bastra-recall work and the brand name is claimed on npm. The real // CLI lives in @bastra-recall/daemon; importing its entry module runs it against // process.argv cli.ts calls main unconditionally, no main-module guard , so // argv and exit codes pass through untouched. Evidence: `packages/bastra-recall/bin/bastra.mjs`
- **Audit Save** (source_file): import { z } from "zod"; import { saveMemory, type SaveMemoryInput, type SaveMemoryResult, } from "./save.js"; import { AuditLog, type AuditEntry, moveToTrash, restoreFromTrash, trashPathFor, } from "./audit-log.js"; import type { Vault } from "./vault.js"; import { access, readFile } from "node:fs/promises"; import matter from "gray-matter"; ⋮---- export type AuditContext = z.infer ; ⋮---- export async function auditedSave args: { vault: Vault; auditLog: AuditLog; vaultRoot: string; input: SaveMemoryInput; context: AuditContext; } : Promise ⋮---- async function readFrontmatter filePath: string, : Promise null ⋮---- async function fileExists p: string : Promise Evidence: `packages/core/src/audit-save.ts`
- **Embeddings** (source_file): import { promises as fs } from "node:fs"; ⋮---- import type { Memory } from "./schema.js"; import type { Vault, VaultEvent } from "./vault.js"; import { EmbedCache, hashEmbedContent } from "./embed-cache.js"; import { assertLocalOrOptIn } from "./ollama-egress.js"; ⋮---- function backpressureLimit : number ⋮---- function backpressureStallMs : number ⋮---- export interface EmbeddingProvider { readonly id: string; readonly dim: number; embed texts: string : Promise ; } ⋮---- embed texts: string : Promise ; ⋮---- export class OpenAIEmbeddingProvider implements EmbeddingProvider ⋮---- constructor opts: { apiKey: string; model?: string; dim?: number; } ⋮---- async embed texts: string : Promise ⋮… Evidence: `packages/core/src/embeddings.ts`
- **Save** (source_file): import { writeFile, access, mkdir, unlink } from "node:fs/promises"; import { join, dirname, resolve, sep } from "node:path"; import { z } from "zod"; import matter from "gray-matter"; import { MemoryTypeEnum, isPathSafeComponent } from "./schema.js"; import { clampSummary, SUMMARY MAX } from "./summary.js"; ⋮---- export type SaveMemoryInput = z.infer ; ⋮---- export interface SaveMemoryResult { id: string; file path: string; created: boolean; summary note?: string; } ⋮---- function isPathSafeFolder value: string : boolean ⋮---- / Marker-Kommentare, zwischen denen der RelatedEnricher die Auto-Wikilink- Section im Body verwaltet. Obsidian rendert HTML-Kommentare als unsichtbar, Wikilinks dazw… Evidence: `packages/core/src/save.ts`
- **Backfill Related** (source_file): import { Vault, EmbeddingIndex, OpenAIEmbeddingProvider, OllamaEmbeddingProvider, RelatedEnricher, extractWikilinks, type EmbeddingProvider, } from "@bastra-recall/core"; import { readFile, writeFile } from "node:fs/promises"; import matter from "gray-matter"; ⋮---- async function main : Promise ⋮---- async function rewriteFrontmatterField filePath: string, field: string, value: unknown, : Promise ⋮---- function sameStringSet a: string , b: string : boolean ⋮---- function sleep ms: number : Promise ⋮---- function pickEmbeddingProvider : EmbeddingProvider null ⋮---- function envInt name: string, fallback: number : number ⋮---- function envFloat name: string, fallback: number : number ⋮---- f… Evidence: `packages/daemon/scripts/backfill-related.ts`
- **Stats** (source_file): import { readdir, readFile } from "node:fs/promises"; import { existsSync } from "node:fs"; import { join } from "node:path"; import { homedir } from "node:os"; ⋮---- function defaultLogDir : string ⋮---- interface AnyEvent { kind: string; ts: string; k: string : unknown; } ⋮---- function pct n: number, total: number : string ⋮---- function median xs: number : number ⋮---- function p95 xs: number : number ⋮---- async function loadEvents : Promise ⋮---- function summarizeHook events: AnyEvent : void ⋮---- function summarizeSessionHook events: AnyEvent : void ⋮---- function summarizeMcp events: AnyEvent : void ⋮---- function summarizeFollowThrough events: AnyEvent : void ⋮---- function summar… Evidence: `packages/daemon/scripts/stats.ts`
- **Curator** (source_file): import { mkdir, readFile, rename, writeFile } from "node:fs/promises"; import { join } from "node:path"; import type { UsageAggregate } from "./usage-sidecar.js"; ⋮---- export interface StaleEntry { stale since: string; reason: string; } ⋮---- export interface CuratorState { last run at?: string; observed since?: string; stale: Record ; } ⋮---- function statePath vaultRoot: string : string ⋮---- export async function loadCuratorState vaultRoot: string : Promise ⋮---- export async function saveCuratorState vaultRoot: string, state: CuratorState : Promise ⋮---- export function shouldRunCurator i: { nowMs: number; lastRunAt?: string; lastActivityMs?: number; intervalMs?: number; minIdleMs?: nu… Evidence: `packages/daemon/src/curator.ts`
- **Http** (source_file): import { createServer, type Server, type IncomingMessage, type ServerResponse } from "node:http"; import { timingSafeEqual } from "node:crypto"; import type { AddressInfo } from "node:net"; import type { Vault, SearchIndex, RecallStage, StageListener, EmbeddingRuntimeHealth } from "@bastra-recall/core"; import { fireAndForget, type Telemetry } from "./telemetry.js"; import { recallHandler, loadMemoryHandler, saveMemoryHandler, toLeanHit, type ToolDeps, } from "./tool-handlers.js"; import { expandQuery, type BridgePool } from "./learned-recall/bridges.js"; import { type SupportedLanguage } from "./learned-recall/language.js"; import { FindDocumentArgs, ReadDocumentArgs, OpenDocumentArgs, fin… Evidence: `packages/daemon/src/http.ts`
- **Index** (source_file): import { Server } from "@modelcontextprotocol/sdk/server/index.js"; import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"; import { CallToolRequestSchema, ListToolsRequestSchema, } from "@modelcontextprotocol/sdk/types.js"; import { Vault, SearchIndex, EmbeddingIndex, OpenAIEmbeddingProvider, OllamaEmbeddingProvider, RelatedEnricher, TriggerExpander, pickPhrase, banterModeFromEnv, progressIndexFor, RECALL STAGE ORDER, type EmbeddingProvider, type RecallStage, type StageListener, } from "@bastra-recall/core"; ⋮---- import { Telemetry, logDirFor } from "./telemetry.js"; import { startHttpServer } from "./http.js"; import { recordUsage } from "./usage-sidecar.js"; im… Evidence: `packages/daemon/src/index.ts`
- **Mcp Forwarder** (source_file): import { Server } from "@modelcontextprotocol/sdk/server/index.js"; import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"; import { CallToolRequestSchema, ListToolsRequestSchema, } from "@modelcontextprotocol/sdk/types.js"; import { spawn } from "node:child process"; ⋮---- import { fileURLToPath } from "node:url"; import { pickPhrase, pickToolPhrase, banterModeFromEnv, progressIndexFor, RECALL STAGE ORDER, type RecallStage, } from "@bastra-recall/core"; import { ALL TOOL DEFS } from "./tool-defs.js"; import { claudeSessionPid, sessionFeedPath, STATUSLINE DIR, reapStaleFeeds } from "./statusline-session.js"; import { commandOf, parentPidOf } from "./reap-forwarders… Evidence: `packages/daemon/src/mcp-forwarder.ts`
- **Index** (source_file): import type { ClaudeHookData } from "./utils/claude"; ⋮---- import process from "node:process"; import { json } from "node:stream/consumers"; import { PowerlineRenderer } from "./powerline"; import { loadConfigFromCLI } from "./config/loader"; import { debug } from "./utils/logger"; ⋮---- function showHelpText : void ⋮---- async function main : Promise Evidence: `packages/statusline/src/index.ts`
- **Index** (source_file): import { darkTheme, darkAnsi256Theme, darkAnsiTheme } from "./dark"; import { lightTheme, lightAnsi256Theme, lightAnsiTheme } from "./light"; import { nordTheme, nordAnsi256Theme, nordAnsiTheme } from "./nord"; import { tokyoNightTheme, tokyoNightAnsi256Theme, tokyoNightAnsiTheme, } from "./tokyo-night"; import { rosePineTheme, rosePineAnsi256Theme, rosePineAnsiTheme, } from "./rose-pine"; import { gruvboxTheme, gruvboxAnsi256Theme, gruvboxAnsiTheme } from "./gruvbox"; ⋮---- export interface SegmentColor { bg: string; fg: string; bold?: boolean; } ⋮---- export interface ColorTheme { directory: SegmentColor; git: SegmentColor; model: SegmentColor; session: SegmentColor; block: SegmentColor;… Evidence: `packages/statusline/src/themes/index.ts`
- **Security Policy** (documentation): Bastra.Recall is pre-1.0. Security fixes target the latest public release and main . Older beta releases may receive fixes only when the patch is small and low-risk. Evidence: `SECURITY.md`
- **Supporters** (documentation): Bastra is built solo. Every sponsor, every star, every issue makes the project sustainable. Thank you. 💖 Evidence: `SUPPORTERS.md`
- **Bastra Recall — Stress Eval Report** (documentation): - Run date: 2026-05-27T15:06:38.294Z - Vault path: ~/Library/CloudStorage/GoogleDrive-nevoigt@n0mad.ai/Meine Ablage/OBSIDIAN/Daniel Nevoigt - Vault size: 163 memories - Search provider: BM25-only Evidence: `packages/daemon/scripts/stress-report.md`
- **Rule** (documentation): Rule Randomize each retry delay within the backoff window. Evidence: `packages/eval/fixtures/eval-vault/memories/backoff-jitter.md`
- **Rule** (documentation): Rule Add min-width: 0 to a flex child that should be allowed to shrink. Evidence: `packages/eval/fixtures/eval-vault/memories/css-flexbox-min-width-zero.md`
- **Rule** (documentation): Rule Pick debounce when you want the final value, throttle when you want a steady cadence. Evidence: `packages/eval/fixtures/eval-vault/memories/debounce-vs-throttle.md`
- **Rule** (documentation): Rule Order Dockerfile steps from least- to most-frequently changing. Evidence: `packages/eval/fixtures/eval-vault/memories/docker-manifest-before-source.md`
- **Rule** (documentation): Rule Style focus, don't delete it; prefer :focus-visible . Evidence: `packages/eval/fixtures/eval-vault/memories/focus-ring-keyboard.md`
- **Rule** (documentation): Rule Make a --fixup commit targeting the original, then autosquash on rebase. Evidence: `packages/eval/fixtures/eval-vault/memories/git-rebase-autosquash-fixup.md`
- **Rule** (documentation): Rule Store amounts as integer cents or a fixed-precision decimal. Evidence: `packages/eval/fixtures/eval-vault/memories/money-not-float.md`
- **Rule** (documentation): Rule Match the predicate's type to the indexed column's type. Evidence: `packages/eval/fixtures/eval-vault/memories/postgres-index-type-mismatch.md`
- **Rule** (documentation): Rule Secrets live in env/secret manager; a committed secret is a compromised secret. Evidence: `packages/eval/fixtures/eval-vault/memories/secrets-in-env.md`
- **Rule** (documentation): Rule UTC at rest and in logic; local zone only at the presentation edge. Evidence: `packages/eval/fixtures/eval-vault/memories/timestamps-store-utc.md`
- **Tsconfig** (structured_config): { "compilerOptions": { "target": "ES2022", "module": "NodeNext", "moduleResolution": "NodeNext", "lib": "ES2022" , "types": "node" , "outDir": "./dist", "rootDir": "./src", "strict": true, "esModuleInterop": true, "skipLibCheck": true, "forceConsistentCasingInFileNames": true, "resolveJsonModule": true, "declaration": true, "declarationMap": true, "sourceMap": true, "noUnusedLocals": true, "noUnusedParameters": false, "noImplicitReturns": true, "noFallthroughCasesInSwitch": true }, "include": "src/ / .ts" , "exclude": "node modules", "dist" } Evidence: `packages/core/tsconfig.json`
- **Tsconfig** (structured_config): { "compilerOptions": { "target": "ES2022", "module": "NodeNext", "moduleResolution": "NodeNext", "lib": "ES2022" , "types": "node" , "outDir": "./dist", "rootDir": "./src", "strict": true, "esModuleInterop": true, "skipLibCheck": true, "forceConsistentCasingInFileNames": true, "resolveJsonModule": true, "declaration": false, "sourceMap": true, "noUnusedLocals": true, "noUnusedParameters": false, "noImplicitReturns": true, "noFallthroughCasesInSwitch": true }, "include": "src/ / .ts" , "exclude": "node modules", "dist", "test-vault" } Evidence: `packages/daemon/tsconfig.json`
- **Cases** (structured_config): { " comment": "Paraphrased held-out queries for the synthetic eval-vault. Each paraphrase is phrased to share as few literal tokens with the memory's recall when as is realistic, simulating how a future session describes the situation in its own words. The 'anti' set is off-vault: no memory should genuinely match.", "paraphrased": { "id": "git-rebase-autosquash-fixup", "label": "tidy history before PR", "paraphrases": "tidy my branch history before opening a pull request", "fold a follow-up edit into an earlier commit automatically", "combine small correction commits into one" }, { "id": "css-flexbox-min-width-zero", "label": "flex item won't shrink", "paraphrases": "long string blows out t… Evidence: `packages/eval/fixtures/cases.json`
- **Personas** (structured_config): { " comment": "Simulated-user queries for the synthetic eval-vault: 10 personas x 10 memories, each persona phrasing every situation in its own voice junior/senior/non-native/terse/verbose/concept-namer/error-message/why-asker/outsider/mobile-typo . LLM-generated on purpose — production memories and recall queries are both AI-mediated, so this is the ecologically faithful test see persona-lift.ts / README . Diversity is guarded: mean pairwise query similarity ~0.12, 10/10 distinct voices. Like cases.json, the synthetic-vault number is a mechanism demo, not a headline.", "personas": { "junior": { "backoff-jitter": "why do all my requests fail at the same time after server comes back", "css-f… Evidence: `packages/eval/fixtures/personas.json`
- The remaining 6 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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/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: `README.md`, `packages/bastra-recall/README.md`, `packages/daemon/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.

- **Overview and Installation (Guided Wizard, Distros, First-Run)**: importance `high`
  - source_paths: README.md, package.json, distribution/Install Bastra.command, distribution/homebrew/bastra-recall.rb, packages/bastra-recall/bin/bastra.mjs
- **System Architecture and Memory Pipeline (Vault, Index, Recall/Save, Schema)**: importance `high`
  - source_paths: docs/architecture.md, docs/memory-schema.md, docs/triggers.md, docs/taxonomy.md, docs/openapi.yaml
- **AI Client Integration, Reflex Layer, and Transport (MCP Forwarder, REST, Hooks)**: importance `high`
  - source_paths: docs/hooks.md, docs/issues/cc-mcp-progress-rendering.md, packages/skill/SKILL.md, packages/daemon/src/index.ts, packages/daemon/src/http.ts
- **Self-Improving Lifecycle, Telemetry, and Operational Concerns**: importance `high`
  - source_paths: PLAN.md, docs/eval-geometry.md, docs/commons.md, docs/survival.md, CHANGELOG.md

## Repo Inspection Evidence

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `7bd3ff57747258bc186898a32008f2d71921505f`
- inspected_files: `README.md`, `package.json`, `docs/architecture.md`, `docs/commons.md`, `docs/eval-geometry.md`, `docs/hooks.md`, `docs/issues/cc-mcp-progress-rendering.md`, `docs/memory-schema.md`, `docs/openapi.yaml`, `docs/survival.md`, `docs/taxonomy.md`, `docs/triggers.md`, `packages/bastra-recall/README.md`, `packages/bastra-recall/package.json`, `packages/core/__tests__/curator-demotion.test.ts`, `packages/core/__tests__/doc-damping.test.ts`, `packages/core/__tests__/embed-backpressure.test.ts`, `packages/core/__tests__/embed-cache.test.ts`, `packages/core/__tests__/embed-health.test.ts`, `packages/core/__tests__/matched-recall-when.test.ts`

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: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: brew install fails on untrusted tap — docs and Install.command miss the new 'brew trust' step
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: brew install fails on untrusted tap — docs and Install.command miss the new 'brew trust' step. Context: Observed during installation or first-run setup.
- Why it matters: Developers may fail before the first successful local run: brew install fails on untrusted tap — docs and Install.command miss the new 'brew trust' step
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/182
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 2: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: homebrew formula build fails — builds only the daemon workspace, core/statusline dist missing from the tarball
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: homebrew formula build fails — builds only the daemon workspace, core/statusline dist missing from the tarball. Context: Observed when using node
- Why it matters: Developers may fail before the first successful local run: homebrew formula build fails — builds only the daemon workspace, core/statusline dist missing from the tarball
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/184
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 3: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: install via npx registers the forwarder path inside the ephemeral npx cache
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: install via npx registers the forwarder path inside the ephemeral npx cache. Context: Observed when using node
- Why it matters: Developers may fail before the first successful local run: install via npx registers the forwarder path inside the ephemeral npx cache
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/180
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 4: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: install: first-run wall — 'vault path required' error instead of a guided vault choice
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: install: first-run wall — 'vault path required' error instead of a guided vault choice. Context: Observed when using node, macos
- Why it matters: Developers may fail before the first successful local run: install: first-run wall — 'vault path required' error instead of a guided vault choice
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/178
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 5: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: test: verify the Ollama fresh-install (brew) path on a clean macOS env
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: test: verify the Ollama fresh-install (brew) path on a clean macOS env. Context: Observed when using node, macos
- Why it matters: Developers may fail before the first successful local run: test: verify the Ollama fresh-install (brew) path on a clean macOS env
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/90
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 6: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: uninstall all: skill kept as 'shared with Claude Desktop' although Desktop was uninstalled in the same run
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: uninstall all: skill kept as 'shared with Claude Desktop' although Desktop was uninstalled in the same run. Context: Observed during installation or first-run setup.
- Why it matters: Developers may fail before the first successful local run: uninstall all: skill kept as 'shared with Claude Desktop' although Desktop was uninstalled in the same run
- Evidence: failure_mode_cluster:github_issue | https://github.com/n0mad-ai/bastra-recall/issues/181
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 7: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: v0.6.0-beta.1
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v0.6.0-beta.1. Context: Observed when using node
- Why it matters: Upgrade or migration may change expected behavior: v0.6.0-beta.1
- Evidence: failure_mode_cluster:github_release | https://github.com/n0mad-ai/bastra-recall/releases/tag/v0.6.0-beta.1
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 8: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: v0.6.5-beta.1
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v0.6.5-beta.1. Context: Observed when using node
- Why it matters: Upgrade or migration may change expected behavior: v0.6.5-beta.1
- Evidence: failure_mode_cluster:github_release | https://github.com/n0mad-ai/bastra-recall/releases/tag/v0.6.5-beta.1
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 9: Installation risk requires verification

- Trigger: Developers should check this installation risk before relying on the project: v0.6.6-beta.1
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v0.6.6-beta.1. Context: Observed when using node
- Why it matters: Upgrade or migration may change expected behavior: v0.6.6-beta.1
- Evidence: failure_mode_cluster:github_release | https://github.com/n0mad-ai/bastra-recall/releases/tag/v0.6.6-beta.1
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.