# memforge - 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 memforge. 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

## What It Can Do

- **Multi-Host Install and Distribution** (Verify after install): The project contains plugin or marketplace configuration, indicating it targets install and distribution across one or more AI hosts. Evidence: `public/ai-plugin.json` 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: `CLAUDE.md`, `README.md` Claim: `clm_0002` supported 0.86

## How to Start

- `pip install memforge` Evidence: `README.md` Claim: `clm_0004` supported 0.86
- `npm install @salishforge/memforge` Evidence: `README.md` Claim: `clm_0005` supported 0.86
- `npm install -g @salishforge/memforge` Evidence: `README.md` Claim: `clm_0006` supported 0.86
- `git clone https://github.com/salishforge/memforge.git` Evidence: `README.md` Claim: `clm_0007` supported 0.86
- `curl -X POST http://localhost:3333/memory/agent-1/add \` Evidence: `README.md` Claim: `clm_0008` supported 0.86
- `curl "http://localhost:3333/memory/agent-1/query?q=user+preferences&mode=hybrid" \` Evidence: `README.md` Claim: `clm_0009` supported 0.86
- `curl -X POST http://localhost:3333/memory/agent-1/sleep \` Evidence: `README.md` Claim: `clm_0010` supported 0.86
- `curl -X POST http://localhost:3333/memory/agent-1/active-recall \` Evidence: `README.md` Claim: `clm_0011` supported 0.86
- `curl "http://localhost:3333/memory/agent-1/query?q=user+preferences" \` Evidence: `CLAUDE.md` Claim: `clm_0009` supported 0.86, `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**: 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: 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
- **Capability exists: Multi-Host Install and Distribution** (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: `public/ai-plugin.json` 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: `CLAUDE.md`, `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_0004` 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`, `public/ai-plugin.json`
- **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. Evidence: `public/ai-plugin.json`
- **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`, `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: `CLAUDE.md`, `public/ai-plugin.json`
- **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`, `README.md`, `public/ai-plugin.json`
- **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: `ARCHITECTURE.md`, `BENCHMARK_IMPLEMENTATION.md`, `CHANGELOG.md`, `CLAUDE.md` et al.
- **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_0013` inferred 0.45
- **Host AI plugin or Skill rule conflicts**: New rules may change how the user's existing host AI behaves. Mitigation: Inspect the plugin manifest and Skill files before installing, and test in isolation if needed. Evidence: `public/ai-plugin.json` Claim: `clm_0014` supported 0.86
- **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`, `README.md` Claim: `clm_0015` 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

- **Multi-Host Install and Distribution**: 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: `public/ai-plugin.json` 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: `CLAUDE.md`, `README.md` Claim: `clm_0002` supported 0.86

### Context Scale

- Total files: 135
- Important-file coverage: 40/135
- Evidence index entries: 79
- Role / Skill entries: 27

### 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 memforge, 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 memforge 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 memforge, 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 27 role / Skill / project-doc entries.

- **MemForge — Project Instructions for AI Agents** (project_doc): MemForge — Project Instructions for AI Agents Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CLAUDE.md`
- **MemForge** (project_doc): ! License: MIT https://img.shields.io/badge/License-MIT-blue.svg LICENSE ! npm version https://img.shields.io/npm/v/@salishforge/memforge.svg?label=npm&color=cb3837 https://www.npmjs.com/package/@salishforge/memforge ! npm downloads https://img.shields.io/npm/dm/@salishforge/memforge.svg?color=blue https://www.npmjs.com/package/@salishforge/memforge ! Status: Beta https://img.shields.io/badge/Status-Beta-yellow.svg… Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `README.md`
- **MemForge Benchmarks** (project_doc): Retrieval quality and performance benchmarks for the MemForge memory system. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `benchmarks/README.md`
- **MemForge Python SDK** (project_doc): Python client for MemForge https://github.com/salishforge/memforge — neuroscience-inspired memory for AI agents. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `python/README.md`
- **LongMemEval QA Accuracy Benchmark** (project_doc): This benchmark runs the full LongMemEval pipeline : retrieve → generate answer → LLM judge. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `benchmarks/longmemeval-qa/README.md`
- **MemForge for Claude Desktop** (project_doc): Persistent long-term memory for Claude Desktop via MCP. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `plugins/claude-desktop/README.md`
- **MemForge for Microsoft 365 Copilot** (project_doc): Persistent long-term memory for M365 Copilot agents via API Plugin or Power Automate. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `plugins/m365-copilot/README.md`
- **Contributing to MemForge** (project_doc): Thank you for your interest in contributing to MemForge. This document explains how to get started, what we look for in contributions, and how the project is organized. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CONTRIBUTING.md`
- **MemForge Adversarial Security Assessment** (project_doc): MemForge Adversarial Security Assessment Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `ADVERSARIAL-ASSESSMENT.md`
- **MemForge — Architecture** (project_doc): MemForge is a neuroscience-inspired memory system for AI agents. This document describes the internal architecture, data models, and processing pipelines. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `ARCHITECTURE.md`
- **Benchmark Implementation Progress** (project_doc): Last updated: 2026-07-27 Status: Phase 1 P0 credibility fixes complete, QA accuracy harness implemented Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `BENCHMARK_IMPLEMENTATION.md`
- **Changelog** (project_doc): All notable changes to MemForge are documented here. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CHANGELOG.md`
- **MemForge × Claude Dreaming — Feature Review & Integration Plan** (project_doc): MemForge × Claude Dreaming — Feature Review & Integration Plan Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CLAUDE_DREAMING_PLAN.md`
- **Contributor Covenant Code of Conduct** (project_doc): Contributor Covenant Code of Conduct Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `CODE_OF_CONDUCT.md`
- **Development Guide** (project_doc): - Node.js = 22 - PostgreSQL 16+ with extensions: - pgvector 0.5+ — vector similarity search; 0.5+ required for halfvec float16 storage - pg trgm — trigram fuzzy matching used for entity dedup and search fallback - Redis 7+ optional — MemForge works without it, just slower - Git Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `DEVELOPMENT.md`
- **Git Commit Commands for Benchmark Plan** (project_doc): Git Commit Commands for Benchmark Plan Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `GIT_COMMIT_COMMANDS.md`
- **MemForge Security Hardening Plan** (project_doc): Based on the Adversarial Security Assessment ADVERSARIAL-ASSESSMENT.md conducted 2026-04-08. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `HARDENING-PLAN.md`
- **Integrating MemForge with AI Agents** (project_doc): Integrating MemForge with AI Agents Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `INTEGRATION.md`
- **MemForge — Competitive Analysis & Next-Gen Recommendations** (project_doc): MemForge — Competitive Analysis & Next-Gen Recommendations Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `NEXTGEN-RECOMMENDATIONS.md`
- **Phase 5: Autonomous Knowledge Architecture — Plan & Funding Brief** (project_doc): Phase 5: Autonomous Knowledge Architecture — Plan & Funding Brief Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `PHASE_5_PLAN.md`
- **MemForge Roadmap** (project_doc): AI agents today are stateless by default. Every conversation starts from zero. The few systems that add memory treat it as a passive store — save facts, retrieve facts, hope for the best. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `ROADMAP.md`
- **Security Policy** (project_doc): Version Supported --------- -------------------- 2.1.x Yes 2.0.x Security fixes only < 2.0 No Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `SECURITY.md`
- **MemForge — Memory Architecture Specification** (project_doc): MemForge — Memory Architecture Specification Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `SPECIFICATION.md`
- **MemForge Threat Model** (project_doc): A comprehensive security analysis of MemForge's architecture, identifying attack vectors, bypass techniques, and integrity risks. This document is intended for security reviewers, contributors, and operators evaluating MemForge for production use. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `THREAT_MODEL.md`
- **MemForge Benchmark Results** (project_doc): ⚠️ RETRACTED — do not cite the figures below Every Recall@k number on this page is invalid. The scorer that produced them called recallAtK ids, answers, ids.length — passing the candidate list's own length as k , so the internal slice 0, k never truncated and R@1, R@3, R@5 and R@10 were all computed over the entire retrieved set. The inflation is not marginal. Consolidation packs many sessions into each warm-tier ro… Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `benchmarks/RESULTS.md`
- **MemForge Tools for Claude** (project_doc): This document describes all memory tools available to Claude when MemForge is connected via MCP. Use these tools to build persistent memory across conversations — remembering user preferences, tracking project context, learning from mistakes, and improving over time. Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `plugins/claude-desktop/TOOLS.md`
- **MemForge Tools for Microsoft 365 Copilot** (project_doc): MemForge Tools for Microsoft 365 Copilot Activation hint: Reference this when the user needs to understand the project's structure, install path, or boundaries. Evidence: `plugins/m365-copilot/TOOLS.md`

## Evidence Index

- Indexed 79 evidence entries.

- **MemForge — Project Instructions for AI Agents** (documentation): MemForge — Project Instructions for AI Agents Evidence: `CLAUDE.md`
- **MemForge** (documentation): ! License: MIT https://img.shields.io/badge/License-MIT-blue.svg LICENSE ! npm version https://img.shields.io/npm/v/@salishforge/memforge.svg?label=npm&color=cb3837 https://www.npmjs.com/package/@salishforge/memforge ! npm downloads https://img.shields.io/npm/dm/@salishforge/memforge.svg?color=blue https://www.npmjs.com/package/@salishforge/memforge ! Status: Beta https://img.shields.io/badge/Status-Beta-yellow.svg project-status ! Node.js https://img.shields.io/badge/Node.js-%3E%3D22-green.svg https://nodejs.org ! TypeScript https://img.shields.io/badge/TypeScript-5.7-blue.svg https://www.typescriptlang.org ! PostgreSQL https://img.shields.io/badge/PostgreSQL-16+-336791.svg https://www.pos… Evidence: `README.md`
- **MemForge Benchmarks** (documentation): Retrieval quality and performance benchmarks for the MemForge memory system. Evidence: `benchmarks/README.md`
- **MemForge Python SDK** (documentation): Python client for MemForge https://github.com/salishforge/memforge — neuroscience-inspired memory for AI agents. Evidence: `python/README.md`
- **LongMemEval QA Accuracy Benchmark** (documentation): This benchmark runs the full LongMemEval pipeline : retrieve → generate answer → LLM judge. Evidence: `benchmarks/longmemeval-qa/README.md`
- **MemForge for Claude Desktop** (documentation): Persistent long-term memory for Claude Desktop via MCP. Evidence: `plugins/claude-desktop/README.md`
- **MemForge for Microsoft 365 Copilot** (documentation): Persistent long-term memory for M365 Copilot agents via API Plugin or Power Automate. Evidence: `plugins/m365-copilot/README.md`
- **Package** (package_manifest): { "name": "@salishforge/memforge", "version": "3.7.0", "description": "Neuroscience-inspired agent memory — sleep cycles, memory revision, knowledge graph, procedural learning, vector search, and MCP integration", "type": "module", "main": "./dist/index.js", "types": "./dist/index.d.ts", "exports": { ".": { "import": "./dist/index.js", "types": "./dist/index.d.ts" }, "./client": { "import": "./dist/client.js", "types": "./dist/client.d.ts" }, "./tools": { "import": "./dist/tool-definitions.js", "types": "./dist/tool-definitions.d.ts" } }, "bin": { "memforge": "./dist/server.js", "memforge-mcp": "./dist/mcp.js" }, "scripts": { "build": "tsc", "start": "node dist/server.js", "dev": "tsx watch… Evidence: `package.json`
- **Contributing to MemForge** (documentation): Thank you for your interest in contributing to MemForge. This document explains how to get started, what we look for in contributions, and how the project is organized. 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`
- **MemForge Adversarial Security Assessment** (documentation): MemForge Adversarial Security Assessment Evidence: `ADVERSARIAL-ASSESSMENT.md`
- **MemForge — Architecture** (documentation): MemForge is a neuroscience-inspired memory system for AI agents. This document describes the internal architecture, data models, and processing pipelines. Evidence: `ARCHITECTURE.md`
- **Benchmark Implementation Progress** (documentation): Last updated: 2026-07-27 Status: Phase 1 P0 credibility fixes complete, QA accuracy harness implemented Evidence: `BENCHMARK_IMPLEMENTATION.md`
- **Changelog** (documentation): All notable changes to MemForge are documented here. Evidence: `CHANGELOG.md`
- **MemForge × Claude Dreaming — Feature Review & Integration Plan** (documentation): MemForge × Claude Dreaming — Feature Review & Integration Plan Evidence: `CLAUDE_DREAMING_PLAN.md`
- **Contributor Covenant Code of Conduct** (documentation): Contributor Covenant Code of Conduct Evidence: `CODE_OF_CONDUCT.md`
- **Development Guide** (documentation): - Node.js = 22 - PostgreSQL 16+ with extensions: - pgvector 0.5+ — vector similarity search; 0.5+ required for halfvec float16 storage - pg trgm — trigram fuzzy matching used for entity dedup and search fallback - Redis 7+ optional — MemForge works without it, just slower - Git Evidence: `DEVELOPMENT.md`
- **Git Commit Commands for Benchmark Plan** (documentation): Git Commit Commands for Benchmark Plan Evidence: `GIT_COMMIT_COMMANDS.md`
- **MemForge Security Hardening Plan** (documentation): Based on the Adversarial Security Assessment ADVERSARIAL-ASSESSMENT.md conducted 2026-04-08. Evidence: `HARDENING-PLAN.md`
- **Integrating MemForge with AI Agents** (documentation): Integrating MemForge with AI Agents Evidence: `INTEGRATION.md`
- **MemForge — Competitive Analysis & Next-Gen Recommendations** (documentation): MemForge — Competitive Analysis & Next-Gen Recommendations Evidence: `NEXTGEN-RECOMMENDATIONS.md`
- **Phase 5: Autonomous Knowledge Architecture — Plan & Funding Brief** (documentation): Phase 5: Autonomous Knowledge Architecture — Plan & Funding Brief Evidence: `PHASE_5_PLAN.md`
- **MemForge Roadmap** (documentation): AI agents today are stateless by default. Every conversation starts from zero. The few systems that add memory treat it as a passive store — save facts, retrieve facts, hope for the best. Evidence: `ROADMAP.md`
- **Security Policy** (documentation): Version Supported --------- -------------------- 2.1.x Yes 2.0.x Security fixes only < 2.0 No Evidence: `SECURITY.md`
- **MemForge — Memory Architecture Specification** (documentation): MemForge — Memory Architecture Specification Evidence: `SPECIFICATION.md`
- **MemForge Threat Model** (documentation): A comprehensive security analysis of MemForge's architecture, identifying attack vectors, bypass techniques, and integrity risks. This document is intended for security reviewers, contributors, and operators evaluating MemForge for production use. Evidence: `THREAT_MODEL.md`
- **MemForge Benchmark Results** (documentation): ⚠️ RETRACTED — do not cite the figures below Every Recall@k number on this page is invalid. The scorer that produced them called recallAtK ids, answers, ids.length — passing the candidate list's own length as k , so the internal slice 0, k never truncated and R@1, R@3, R@5 and R@10 were all computed over the entire retrieved set. The inflation is not marginal. Consolidation packs many sessions into each warm-tier row, so "the top 5 rows" could hold hundreds of sessions. What was published as "93.2% R@5" actually means "a gold session appeared anywhere among all sessions inside the top 5 rows." That is not LongMemEval's R@5. Fixed in benchmarks/lib/metrics.ts , which now computes two clearly… Evidence: `benchmarks/RESULTS.md`
- **MemForge Tools for Claude** (documentation): This document describes all memory tools available to Claude when MemForge is connected via MCP. Use these tools to build persistent memory across conversations — remembering user preferences, tracking project context, learning from mistakes, and improving over time. Evidence: `plugins/claude-desktop/TOOLS.md`
- **MemForge Tools for Microsoft 365 Copilot** (documentation): MemForge Tools for Microsoft 365 Copilot Evidence: `plugins/m365-copilot/TOOLS.md`
- **Settings.Local** (structured_config): { "permissions": { "allow": "Bash grep -E \"\\\\.\\ ts js json yaml yml md\\ $\" ", "Bash gh api: ", "WebFetch domain:github.com ", "WebSearch", "mcp claude ai GitHub - Salishforge get file contents", "mcp claude ai GitHub - Salishforge search repositories", "Bash npx tsc: ", "Bash npm run: ", "Bash npm install: ", "Bash ./node modules/.bin/tsc --noEmit ", "Bash ./node modules/.bin/tsc ", "Bash wc -l src/ .ts schema/ .sql ", "Bash wc -l src/ .ts " } } Evidence: `.claude/settings.local.json`
- **Tsconfig** (structured_config): { "extends": "../tsconfig.json", "compilerOptions": { "rootDir": "..", "noEmit": true }, "include": " / .ts", "../src/ / .ts" } Evidence: `benchmarks/tsconfig.json`
- **Funding** (structured_config): { "$schema": "https://fundingjson.org/schema/v1.1.0/funding.schema.json", "entity": { "type": "individual", "role": "maintainer", "name": "John Brooke", "email": "john@salishforge.com", "description": "Solo maintainer building an open-source, MIT-licensed stack for long-running, multi-agent AI systems. The thesis is that frontier model labs are rightly focused on raw capability, but LLMs need a substrate that accumulates experience and revises beliefs over time to operate on human timelines. I'm building that substrate \u2014 security-first, zero-trust from day one \u2014 out of pocket on a modest VPS, with two autonomous agents running in production on the stack. Shipping weekly. Based in… Evidence: `funding.json`
- **Glama** (structured_config): { "$schema": "https://glama.ai/mcp/schemas/server.json", "maintainers": "salishforge" } Evidence: `glama.json`
- **Ai Plugin** (structured_config): { "schema version": "v1", "name for human": "MemForge Memory", "name for model": "memforge", "description for human": "Long-term memory for AI agents — store, search, and manage knowledge", "description for model": "MemForge provides persistent memory for AI agents. Use memforge tools to store memories, search past knowledge, and manage an agent's accumulated experience.", "auth": { "type": "service http", "authorization type": "bearer" }, "api": { "type": "openapi", "url": "/api/spec.json" } } Evidence: `public/ai-plugin.json`
- **Server** (structured_config): { "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json", "name": "io.github.salishforge/memforge-mcp", "title": "MemForge", "description": "Neuroscience-inspired memory system for AI agents. Tiered storage hot/warm/cold , hybrid retrieval FTS + pgvector , knowledge graph, and a 10-phase sleep cycle that actively revises and improves stored knowledge during idle periods. 17 MCP tools.", "version": "3.0.0", "repository": { "url": "https://github.com/salishforge/memforge", "source": "github" }, "packages": { "registryType": "npm", "registryBaseUrl": "https://registry.npmjs.org", "identifier": "memforge-mcp", "version": "3.0.0", "runtimeHint": "npx", "transpo… Evidence: `server.json`
- **Tsconfig** (structured_config): { "compilerOptions": { "target": "ES2022", "module": "NodeNext", "moduleResolution": "NodeNext", "outDir": "./dist", "rootDir": "./src", "declaration": true, "declarationMap": true, "sourceMap": true, "strict": true, "noUncheckedIndexedAccess": true, "noImplicitReturns": true, "noFallthroughCasesInSwitch": true, "verbatimModuleSyntax": true, "esModuleInterop": true, "skipLibCheck": true }, "include": "src/ / " , "exclude": "node modules", "dist" } Evidence: `tsconfig.json`
- **Claude Desktop Config** (structured_config): { "mcpServers": { "memforge": { "command": "npx", "args": "memforge-mcp" , "env": { "MEMFORGE URL": "http://localhost:3333", "MEMFORGE TOKEN": "" } } } } Evidence: `plugins/claude-desktop/claude_desktop_config.json`
- **Manifest** (structured_config): { "$schema": "https://developer.microsoft.com/json-schemas/copilot/plugin/v2.2/schema.json", "schema version": "v2.2", "name for human": "MemForge Memory", "description for human": "Long-term memory for AI agents — store, search, and manage knowledge that persists across sessions", "description for model": "MemForge provides persistent memory for AI agents. Use these tools to: store new memories facts, decisions, preferences, errors , search past knowledge with semantic understanding, get relevant context before tasks, and manage knowledge quality through consolidation and reflection. Memory persists across sessions and improves over time through active knowledge management.", "logo url": "… Evidence: `plugins/m365-copilot/manifest.json`
- **Meeting Context** (structured_config): { "name": "MemForge: Pre-Meeting Context", "description": "5 minutes before a calendar event, retrieve relevant memories and send a Teams notification", "trigger": { "type": "Office365Outlook.CalendarEventSoon", "inputs": { "minutesBefore": 5 } }, "actions": { "name": "SearchMemories", "type": "Http", "inputs": { "method": "GET", "uri": "@{parameters 'MemForgeUrl' }/memory/@{parameters 'AgentId' }/query?q=@{encodeUriComponent triggerOutputs ? 'body/subject' }&mode=hybrid&max tokens=2000&limit=5", "headers": { "Authorization": "Bearer @{parameters 'MemForgeToken' }" } } }, { "name": "SendTeamsCard", "type": "MicrosoftTeams.PostMessageToConversation", "runAfter": "SearchMemories", "inputs": {… Evidence: `plugins/power-automate/meeting-context.json`
- **Nightly Consolidation** (structured_config): { "name": "MemForge: Nightly Consolidation & Sleep Cycle", "description": "Run consolidation and sleep cycle at 2:00 AM daily to maintain memory quality", "trigger": { "type": "Recurrence", "inputs": { "frequency": "Day", "interval": 1, "startTime": "2026-01-01T02:00:00Z", "timeZone": "UTC" } }, "actions": { "name": "Consolidate", "type": "Http", "inputs": { "method": "POST", "uri": "@{parameters 'MemForgeUrl' }/memory/@{parameters 'AgentId' }/consolidate", "headers": { "Authorization": "Bearer @{parameters 'MemForgeToken' }", "Content-Type": "application/json" }, "body": {} } }, { "name": "SleepCycle", "type": "Http", "runAfter": "Consolidate", "inputs": { "method": "POST", "uri": "@{param… Evidence: `plugins/power-automate/nightly-consolidation.json`
- **Store Emails** (structured_config): { "name": "MemForge: Store Important Emails", "description": "When an email tagged 'Important' arrives, store it as a memory in MemForge", "trigger": { "type": "Office365Outlook.OnNewEmail", "inputs": { "importance": "High", "includeAttachments": false } }, "actions": { "type": "Http", "inputs": { "method": "POST", "uri": "@{parameters 'MemForgeUrl' }/memory/@{parameters 'AgentId' }/add", "headers": { "Authorization": "Bearer @{parameters 'MemForgeToken' }", "Content-Type": "application/json" }, "body": { "content": "Email from @{triggerOutputs ? 'body/from' } — Subject: @{triggerOutputs ? 'body/subject' } — @{substring triggerOutputs ? 'body/bodyPreview' , 0, min length triggerOutputs ? 'b… Evidence: `plugins/power-automate/store-emails.json`
- **.dockerignore** (source_file): node modules dist .git .github .claude .env .env.local .env.production .env !.env.example !.env.docker tests coverage .md !README.md .log .DS Store Thumbs.db .editorconfig .nvmrc .eslintrc.json tsconfig.json Evidence: `.dockerignore`
- **.editorconfig** (source_file): indent style = space indent size = 2 end of line = lf charset = utf-8 trim trailing whitespace = true insert final newline = true Evidence: `.editorconfig`
- **Docker Compose environment — copy to .env and set POSTGRES PASSWORD** (source_file): Docker Compose environment — copy to .env and set POSTGRES PASSWORD POSTGRES DB=memforge POSTGRES USER=memforge POSTGRES PASSWORD=changeme POSTGRES PORT=5432 REDIS PORT=6379 PORT=3333 Evidence: `.env.docker`
- **─── MemForge Standalone — Environment Configuration ────────────────────────** (source_file): ─── MemForge Standalone — Environment Configuration ──────────────────────── Evidence: `.env.example`
- **Dependencies** (source_file): Environment files contain secrets .env .env.local .env.production .env Evidence: `.gitignore`
- **.nvmrc** (source_file): 22 Evidence: `.nvmrc`
- **─── Build stage ─────────────────────────────────────────────────────────────** (source_file): ─── Build stage ───────────────────────────────────────────────────────────── FROM node:22-alpine AS build Evidence: `Dockerfile`
- **Install Node.js 22 and build tools** (source_file): Install Node.js 22 and build tools RUN apk add --no-cache nodejs npm curl Evidence: `Dockerfile.standalone`
- **Demo** (source_file): set -euo pipefail RESET='\033 0m' BOLD='\033 1m' GREEN='\033 0;32m' YELLOW='\033 1;33m' CYAN='\033 0;36m' RED='\033 0;31m' DIM='\033 2m' ok { printf "${GREEN}${BOLD} ✓${RESET} %s\n" "$ "; } hdr { printf "\n${YELLOW}${BOLD}══ %s${RESET}\n" "$ "; } info { printf "${CYAN} %s${RESET}\n" "$ "; } dim { printf "${DIM} %s${RESET}\n" "$ "; } err { printf "${RED}${BOLD} ✗${RESET} %s\n" "$ " &2; } note { printf " %s\n" "$ "; } clear 2 /dev/null true printf "${BOLD}" printf " ╔══════════════════════════════════════════════════════╗\n" printf " ║ M E M F O R G E — D E M O ║\n" printf " ║ memories that improve while the agent sleeps ║\n" printf " ╚══════════════════════════════════════════════════════╝\n… Evidence: `demo.sh`
- **Docker Compose** (source_file): version: "3.9" services: postgres: image: postgres:16-alpine restart: unless-stopped environment: POSTGRES DB: ${POSTGRES DB:-memforge} POSTGRES USER: ${POSTGRES USER:-memforge} POSTGRES PASSWORD: ${POSTGRES PASSWORD:?Set POSTGRES PASSWORD in .env or environment} volumes: - pgdata:/var/lib/postgresql/data - ./schema/schema.sql:/docker-entrypoint-initdb.d/schema.sql:ro ports: - "${POSTGRES PORT:-5432}:5432" healthcheck: test: "CMD-SHELL", "pg isready -U ${POSTGRES USER:-memforge} -d ${POSTGRES DB:-memforge}" interval: 10s timeout: 5s retries: 5 redis: image: redis:7-alpine restart: unless-stopped ports: - "${REDIS PORT:-6379}:6379" healthcheck: test: "CMD", "redis-cli", "ping" interval: 10s… Evidence: `docker-compose.yml`
- **Docker Entrypoint Standalone** (source_file): set -e PGDATA="${PGDATA:-/var/lib/postgresql/data}" if ! -f "$PGDATA/PG VERSION" ; then echo " standalone Initialising PostgreSQL data directory..." su-exec postgres initdb -D "$PGDATA" --encoding=UTF8 --auth=trust fi echo " standalone Starting PostgreSQL..." su-exec postgres pg ctl -D "$PGDATA" -l "$PGDATA/pg.log" start -w \ -o "-c listen addresses=localhost -c max connections=50" echo " standalone Creating database..." su-exec postgres createdb memforge 2 /dev/null true echo " standalone Applying schema..." su-exec postgres psql -d memforge -f /app/schema/schema.sql 2 /dev/null true for f in $ ls /app/schema/migration- .sql 2 /dev/null sort ; do echo " standalone Applying migration: $f" s… Evidence: `docker-entrypoint-standalone.sh`
- **Claude Tools** (source_file): AGENT ID = "claude-tools-demo" SYSTEM PROMPT = ⋮---- async def dispatch tool client: MemForgeClient, name: str, args: dict str, Any - str ⋮---- r = await client.add args "agent id" , args "content" , ⋮---- results = await client.query args "agent id" , q=args "q" , ⋮---- r = await client.consolidate args "agent id" , args.get "mode" ⋮---- r = await client.active recall args "agent id" , args "context" , args.get "limit", 5 ⋮---- s = await client.stats args "agent id" ⋮---- async def run agent turn client: MemForgeClient, ac: Any, user message: str - str ⋮---- tools = anthropic tools messages: list dict str, Any = {"role": "user", "content": user message} ⋮---- resp = await ac.messages.creat… Evidence: `examples/claude_tools.py`
- **Langchain Memory** (source_file): AGENT ID = "langchain-demo" ⋮---- def build memforge memory agent id: str - "MemForgeMemory" ⋮---- class MemForgeMemory BaseChatMemory ⋮---- memory key: str = "history" agent id: str = "langchain-agent" memforge: Any = None ⋮---- class Config ⋮---- arbitrary types allowed = True ⋮---- def init self, agent id: str = "langchain-agent", kwargs: Any - None ⋮---- @property def memory variables self - list str ⋮---- def load memory variables self, inputs: dict str, Any - dict str, str ⋮---- human input = inputs.get "input", inputs.get "human input", "" context = asyncio.get event loop .run until complete ⋮---- def save context self, inputs: dict str, Any , outputs: dict str, str - None ⋮---- """P… Evidence: `examples/langchain_memory.py`
- **Openai Tools** (source_file): AGENT ID = "openai-tools-demo" SYSTEM PROMPT = ⋮---- async def dispatch tool client: MemForgeClient, name: str, args: dict str, Any - str ⋮---- r = await client.add args "agent id" , args "content" , ⋮---- results = await client.query args "agent id" , q=args "q" , ⋮---- r = await client.consolidate args "agent id" , args.get "mode" ⋮---- r = await client.active recall args "agent id" , args "context" , args.get "limit", 5 ⋮---- s = await client.stats args "agent id" ⋮---- async def run agent turn client: MemForgeClient, oc: Any, user message: str - str ⋮---- tools = openai tools messages: list dict str, Any = ⋮---- resp = await oc.chat.completions.create model="gpt-4o-mini", messages=messa… Evidence: `examples/openai_tools.py`
- **4. Keyword query** (source_file): AGENT = "quickstart-agent-py" ⋮---- async def main - None ⋮---- base url = os.environ.get "MEMFORGE URL", "http://localhost:3333" token = os.environ.get "MEMFORGE TOKEN" ⋮---- result = await client.consolidate AGENT ⋮---- recall = await client.active recall AGENT, "preparing to deploy to production" memories = recall.get "memories", if recall else ⋮---- snippet = str m.get "content", "" :80 relevance = m.get "relevance", "?" ⋮---- 4. Keyword query ⋮---- results = await client.query AGENT, q="payments API rate limit", mode="keyword" ⋮---- 5. Timeline: chronological view ⋮---- timeline = await client.timeline AGENT, limit=3 ⋮---- content = str entry.get "content", "" :80 ⋮---- 6. Memory stati… Evidence: `examples/quickstart.py`
- **Quickstart** (source_file): import { ResilientMemForgeClient } from '../src/client.js'; ⋮---- async function main Evidence: `examples/quickstart.ts`
- **2. Act — call the LLM** (source_file): AGENT ID = "simple-chatbot" SYSTEM PROMPT = "You are a helpful assistant with persistent memory across sessions." ⋮---- context = await memory.get context user message, max tokens=1500, limit=8 ⋮---- system = SYSTEM PROMPT ⋮---- 2. Act — call the LLM response = await openai client.chat.completions.create assistant reply: str = response.choices 0 .message.content or "" ⋮---- 3. Store — record both turns ⋮---- async def main - None ⋮---- openai client = openai.AsyncOpenAI ⋮---- session context = await memory.start session ⋮---- user input = input "You: " .strip ⋮---- reply = await chat turn memory, user input, openai client ⋮---- End of session: consolidate hot-tier events into searchable mem… Evidence: `examples/simple_chatbot.py`
- **Pyproject** (source_file): build-system requires = "hatchling" build-backend = "hatchling.build" Evidence: `python/pyproject.toml`
- **Hnsw Indexes.Example** (source_file): ALTER TABLE warm tier ALTER COLUMN embedding TYPE halfvec 384 USING embedding::halfvec 384 ; CREATE INDEX IF NOT EXISTS warm tier embedding idx ON warm tier USING hnsw embedding halfvec cosine ops ; ALTER TABLE shared memories ALTER COLUMN embedding TYPE halfvec 384 USING embedding::halfvec 384 ; CREATE INDEX IF NOT EXISTS shared memories embedding idx ON shared memories USING hnsw embedding halfvec cosine ops ; Evidence: `schema/hnsw-indexes.example.sql`
- The remaining 19 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`, `benchmarks/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`, `benchmarks/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.

- **MemForge Overview & Core Concepts**: importance `high`
  - source_paths: README.md, SPECIFICATION.md, ROADMAP.md, ARCHITECTURE.md, CLAUDE.md
- **Memory Architecture, Schema & Sleep Cycles**: importance `high`
  - source_paths: src/memory-manager.ts, src/sleep-cycle.ts, src/db.ts, src/schemas.ts, src/embedding.ts
- **REST API, TypeScript & Python SDKs, MCP & Plugins**: importance `high`
  - source_paths: src/server.ts, src/app.ts, src/mcp.ts, src/tool-definitions.ts, src/openapi.ts
- **Deployment, Operations, Security & Known Issues**: importance `high`
  - source_paths: src/auth.ts, src/audit.ts, src/cache.ts, src/config.ts, src/llm-safety.ts

## Repo Inspection Evidence

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `6863fd780c9e990b0f023dae4dd3ba41564e6982`
- inspected_files: `Dockerfile`, `README.md`, `docker-compose.yml`, `package.json`, `examples/claude_tools.py`, `examples/langchain_memory.py`, `examples/openai_tools.py`, `examples/quickstart.py`, `examples/quickstart.ts`, `examples/simple_chatbot.py`, `src/app.ts`, `src/audit.ts`, `src/auth.ts`, `src/cache.ts`, `src/classifier.ts`, `src/client.ts`, `src/config.ts`, `src/dashboard.ts`, `src/db.ts`, `src/dream-runs.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: v3.0.0-beta.3
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v3.0.0-beta.3. Context: Observed when using python
- Why it matters: Upgrade or migration may change expected behavior: v3.0.0-beta.3
- Evidence: failure_mode_cluster:github_release | https://github.com/salishforge/memforge/releases/tag/v3.0.0-beta.3
- 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: v3.0.0-beta.4
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v3.0.0-beta.4. Context: Observed when using node, python
- Why it matters: Upgrade or migration may change expected behavior: v3.0.0-beta.4
- Evidence: failure_mode_cluster:github_release | https://github.com/salishforge/memforge/releases/tag/v3.0.0-beta.4
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 3: Configuration risk requires verification

- Trigger: Developers should check this configuration risk before relying on the project: MemForge v0.1.0-alpha
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: MemForge v0.1.0-alpha. Context: Observed when using docker
- Why it matters: Upgrade or migration may change expected behavior: MemForge v0.1.0-alpha
- Evidence: failure_mode_cluster:github_release | https://github.com/salishforge/memforge/releases/tag/v0.1.0-alpha
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 4: Configuration risk requires verification

- Trigger: Developers should check this configuration risk before relying on the project: v3.0.0-beta.2 — CI Green, Shared Memory, Full Test Coverage
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: v3.0.0-beta.2 — CI Green, Shared Memory, Full Test Coverage. Context: Observed when using node, python, docker
- Why it matters: Upgrade or migration may change expected behavior: v3.0.0-beta.2 — CI Green, Shared Memory, Full Test Coverage
- Evidence: failure_mode_cluster:github_release | https://github.com/salishforge/memforge/releases/tag/v3.0.0-beta.2
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 5: 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/salishforge/memforge
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

### Constraint 6: 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/salishforge/memforge
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.