# plur - 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 plur. 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_0004` supported 0.86
- **Users who want to bring professional workflows into a host AI**: The repo contains Skill documents. Evidence: `packages/claw/clawhub/SKILL.md`, `packages/mcp/packs/effective-memory/SKILL.md`, `skills/plur-memory/SKILL.md`, `skills/plur-session-end/SKILL.md` Claim: `clm_0005` 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/claw/clawhub/SKILL.md`, `packages/mcp/packs/effective-memory/SKILL.md`, `skills/plur-memory/SKILL.md`, `skills/plur-session-end/SKILL.md` Claim: `clm_0001` supported 0.86
- **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: `.claude-plugin/plugin.json`, `packages/claw/openclaw.plugin.json` Claim: `clm_0002` 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`, `packages/claw/README.md`, `packages/cli/README.md` et al. Claim: `clm_0003` supported 0.86

## How to Start

- `npx @plur-ai/mcp init` Evidence: `README.md` Claim: `clm_0006` supported 0.86
- `npx @plur-ai/cli init --domain myapp --scope project:my-app` Evidence: `README.md` Claim: `clm_0007` supported 0.86
- `npm install -g @plur-ai/mcp` Evidence: `README.md` Claim: `clm_0008` supported 0.86
- `pip install plur-hermes` Evidence: `README.md` Claim: `clm_0009` supported 0.86
- `npm install -g @plur-ai/cli` Evidence: `README.md` Claim: `clm_0010` supported 0.86, `clm_0012` supported 0.86
- `pip install plur-ai` Evidence: `README.md` Claim: `clm_0011` supported 0.86
- `npm install -g @plur-ai/cli   # bridge (required)` Evidence: `README.md` Claim: `clm_0012` supported 0.86
- `git clone https://github.com/plur-ai/plur.git` Evidence: `README.md` Claim: `clm_0013` supported 0.86
- `npx @plur-ai/claw setup` Evidence: `packages/claw/README.md` Claim: `clm_0014` supported 0.86, `clm_0016` supported 0.86
- `npx @plur-ai/claw doctor          # read-only: reports which step is failing` Evidence: `packages/claw/README.md` Claim: `clm_0015` 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_0004` 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/claw/clawhub/SKILL.md`, `packages/mcp/packs/effective-memory/SKILL.md`, `skills/plur-memory/SKILL.md`, `skills/plur-session-end/SKILL.md` Claim: `clm_0005` 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/claw/clawhub/SKILL.md`, `packages/mcp/packs/effective-memory/SKILL.md`, `skills/plur-memory/SKILL.md`, `skills/plur-session-end/SKILL.md` Claim: `clm_0001` 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: `.claude-plugin/plugin.json`, `packages/claw/openclaw.plugin.json` Claim: `clm_0002` supported 0.86
- **Capability exists: Command-Line Startup or Install Flow** (supported): You can trust that the project contains signals of this capability; whether it fits your specific task still needs trial or after-install verification. Evidence: `CLAUDE.md`, `README.md`, `packages/claw/README.md`, `packages/cli/README.md` et al. Claim: `clm_0003` 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_0006` 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: `.claude-plugin/plugin.json`, `CLAUDE.md`, `packages/claw/clawhub/SKILL.md`, `packages/claw/openclaw.plugin.json` et al.
- **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: `.claude-plugin/plugin.json`, `packages/claw/openclaw.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.
- **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: `CLAUDE.md`, `README.md`, `packages/claw/README.md`, `packages/cli/README.md` et al.
- **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-plugin/plugin.json`, `CLAUDE.md`, `packages/claw/clawhub/SKILL.md`, `packages/claw/openclaw.plugin.json` et al.
- **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-plugin/plugin.json`, `CLAUDE.md`, `README.md`, `packages/claw/README.md` et al.
- **Environment variables / API keys**: Project entry docs explicitly showing API key, token, secret, or account credential configuration. Why: If a real install needs credentials, use test credentials first and go through a permission/compliance review. Evidence: `CLAUDE.md`, `docs/test-pyramid.md`, `packages/hermes/test/test_learner.py`
- **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_0022` 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: `.claude-plugin/plugin.json`, `packages/claw/openclaw.plugin.json` Claim: `clm_0023` 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`, `packages/claw/README.md`, `packages/cli/README.md` et al. Claim: `clm_0024` 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/claw/clawhub/SKILL.md`, `packages/mcp/packs/effective-memory/SKILL.md`, `skills/plur-memory/SKILL.md`, `skills/plur-session-end/SKILL.md` Claim: `clm_0001` supported 0.86
- **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: `.claude-plugin/plugin.json`, `packages/claw/openclaw.plugin.json` Claim: `clm_0002` 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`, `packages/claw/README.md`, `packages/cli/README.md` et al. Claim: `clm_0003` supported 0.86

### Context Scale

- Total files: 488
- Important-file coverage: 40/488
- Evidence index entries: 76
- Role / Skill entries: 4

### 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 plur, 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 plur 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 plur, 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 4 role / Skill / project-doc entries.

- **plur-memory** (skill): Your memory stays on your machine. No cloud, no tracking, no API key. PLUR makes your OpenClaw remember — and shares that memory with every other tool you use. Activation hint: When the user's task is highly relevant to the workflow described by “plur-memory”, use it for a pre-install experience first, then decide whether to install. Evidence: `packages/claw/clawhub/SKILL.md`
- **Effective Memory** (skill): The essential habits for an AI agent with memory — session bookends, learning triggers, verification, safety, and the operational discipline that turns raw recall into compounding intelligence. Pinned, always-injected. Activation hint: When the user's task is highly relevant to the workflow described by “Effective Memory”, use it for a pre-install experience first, then decide whether to install. Evidence: `packages/mcp/packs/effective-memory/SKILL.md`
- **plur-memory** (skill): Persistent learning for AI agents. Open engram format. Your agent learns from corrections, remembers across sessions, and transfers knowledge across domains. Activation hint: When the user's task is highly relevant to the workflow described by “plur-memory”, use it for a pre-install experience first, then decide whether to install. Evidence: `skills/plur-memory/SKILL.md`
- **plur-session-end** (skill): Extract durable learnings at the end of a session. Saves corrections, preferences, and codebase patterns as engrams — nothing ephemeral, nothing sensitive. Activation hint: When the user's task is highly relevant to the workflow described by “plur-session-end”, use it for a pre-install experience first, then decide whether to install. Evidence: `skills/plur-session-end/SKILL.md`

## Evidence Index

- Indexed 76 evidence entries.

- **Architecture Decision Records** (documentation): ADRs document significant architectural choices — why a decision was made, what alternatives were rejected, and what the trade-offs are. Evidence: `docs/adr/README.md`
- **PLUR for Claude Code** (documentation): Give Claude Code persistent memory. Corrections, preferences, and conventions persist across sessions — stored as plain YAML on your disk, with zero-cost local search. Evidence: `.claude-plugin/README.md`
- **PLUR for OpenClaw** (documentation): Shared memory across your OpenClaw agents. PLUR plugs into the OpenClaw ContextEngine — auto-injecting relevant memories on session start and auto-extracting learnings from conversations. Local-first, no cloud required. Evidence: `.openclaw-plugin/README.md`
- **CLAUDE.md** (documentation): Persistent memory for AI agents. An agent corrected on Monday remembers on Tuesday — across sessions, tools, and machines. Evidence: `CLAUDE.md`
- **PLUR — Your agents share the same memory** (documentation): PLUR — Your agents share the same memory Evidence: `README.md`
- **PLUR — Own Your Intelligence** (documentation): Memory for AI agents as an open, portable, local format you control — not rows in someone else's cloud database. Evidence: `comparisons/README.md`
- **PLUR examples** (documentation): Runnable, copy-pasteable getting-started scripts for @plur-ai/core ../packages/core . Each is self-contained, writes to a throwaway temp store never your real ~/.plur , and prints what it does. Evidence: `examples/README.md`
- **@plur-ai/claw** (documentation): Engram exchange layer for OpenClaw agents. Local-first data control. No cloud routing, no API key, data never leaves your machine. Evidence: `packages/claw/README.md`
- **@plur-ai/cli** (documentation): Persistent memory for AI agents — from the command line. Evidence: `packages/cli/README.md`
- **@plur-ai/core** (documentation): The engine behind PLUR https://plur.ai — persistent memory for AI agents. Evidence: `packages/core/README.md`
- **plur-hermes** (documentation): Persistent memory plugin for Hermes Agent https://github.com/NousResearch/hermes-agent . Your agent corrected on Monday remembers on Tuesday. Evidence: `packages/hermes/README.md`
- **plur-langchain** (documentation): LangChain memory adapter for PLUR https://plur.ai — plug persistent, local-first engram memory into any LangChain chain. Evidence: `packages/langchain/README.md`
- **@plur-ai/mcp** (documentation): Give your AI agent persistent memory. One line in your MCP config — corrections, preferences, and conventions persist across sessions. No workflow changes, no cloud, no API costs for search. Evidence: `packages/mcp/README.md`
- **plur-ai** (documentation): Local-first shared memory for AI agents — as a Python SDK. Evidence: `packages/python/README.md`
- **Package** (package_manifest): { "name": "@plur-ai/examples", "version": "0.0.0", "private": true, "type": "module", "description": "Runnable getting-started examples for @plur-ai/core", "scripts": { "ex:basic": "tsx 01-basic-store-recall.ts", "ex:mcp": "tsx 02-mcp-integration.ts", "ex:scope": "tsx 03-scope-filtering.ts", "all": "pnpm ex:basic && pnpm ex:mcp && pnpm ex:scope" }, "dependencies": { "@plur-ai/core": "workspace: " }, "devDependencies": { "tsx": "^4.22.4" } } Evidence: `examples/package.json`
- **Package** (package_manifest): { "name": "plur", "private": true, "scripts": { "build": "pnpm -r build", "test": "vitest run", "test:integration": "vitest run packages/core/test/remote-integration.test.ts", "test:smoke": "vitest run packages/core/test/remote-smoke.test.ts", "test:watch": "vitest", "bench:micro": "npx tsx benchmark/micro.ts" }, "devDependencies": { "tsup": "^8.5.1", "typescript": "^6.0.2", "vitest": "^4.1.1" }, "pnpm": { "overrides": { "fast-xml-parser": " =5.7.0", "@hono/node-server": " =1.19.13", "openclaw": " =2026.6.5", "undici": " =7.28.0", "linkify-it": " =5.0.1", "js-yaml": " =4.2.0 =7.5.6", "hono": " =4.12.18", "brace-expansion": " =2.0.2", "postcss": " =8.5.14", "yaml": " =2.9.0", "minimatch": "… Evidence: `package.json`
- **Contributing to PLUR** (documentation): Thanks for helping build PLUR — persistent, composable memory for AI agents. This monorepo holds seven packages: @plur-ai/core the engine , @plur-ai/mcp the MCP server , @plur-ai/claw the OpenClaw plugin , @plur-ai/cli the CLI , plur-hermes Hermes Agent plugin , plur-ai Python SDK , and plur-langchain LangChain adapter . For the release process see RELEASING.md RELEASING.md ; for agent/automation working rules see CLAUDE.md CLAUDE.md . Evidence: `CONTRIBUTING.md`
- **Package** (package_manifest): { "name": "@plur-ai/claw", "version": "0.10.0", "type": "module", "main": "dist/index.js", "types": "dist/index.d.ts", "bin": { "claw": "dist/cli.js" }, "files": "dist", "openclaw.plugin.json" , "openclaw": { "extensions": "dist/index.js" , "compat": { "pluginApi": " =2026.3.0" }, "build": { "openclawVersion": "2026.3.7" } }, "scripts": { "build": "tsup", "test": "vitest run", "postinstall": "node dist/setup.js 2 /dev/null true" }, "dependencies": { "@plur-ai/core": "workspace: " }, "peerDependencies": { "openclaw": " =2026.6.5" }, "peerDependenciesMeta": { "openclaw": { "optional": true } }, "devDependencies": { "@types/node": "^22.0.0" }, "license": "Apache-2.0", "description": "PLUR memo… Evidence: `packages/claw/package.json`
- **Package** (package_manifest): { "name": "@plur-ai/cli", "version": "0.14.0", "type": "module", "bin": { "plur": "dist/index.js" }, "main": "dist/index.js", "files": "dist" , "scripts": { "build": "tsup", "test": "vitest run" }, "dependencies": { "@plur-ai/core": "workspace: " }, "devDependencies": { "@types/node": "^25.5.0", "better-sqlite3": "^11.0.0" }, "peerDependencies": { "@plur-ai/core": " =0.3.0" }, "peerDependenciesMeta": { "@plur-ai/core": { "optional": false } }, "license": "Apache-2.0", "description": "PLUR CLI — persistent memory for AI agents from the command line", "keywords": "ai", "memory", "cli", "engram", "agent", "plur" , "homepage": "https://plur.ai", "repository": { "type": "git", "url": "https://gi… Evidence: `packages/cli/package.json`
- **Package** (package_manifest): { "name": "@plur-ai/core", "version": "0.14.0", "type": "module", "main": "dist/index.js", "types": "dist/index.d.ts", "files": "dist" , "scripts": { "build": "tsup", "test": "vitest run", "gen:schemas": "tsx scripts/gen-spec-schemas.ts" }, "dependencies": { "@electric-sql/pglite": "^0.4.6", "js-yaml": "^4.1.0", "zod": "^3.23.0" }, "optionalDependencies": { "@huggingface/transformers": "^3.8.1", "better-sqlite3": "^11.0.0" }, "devDependencies": { "@types/better-sqlite3": "^7.6.0", "@types/js-yaml": "^4.0.0", "@types/node": "^25.5.0", "tsx": "^4.22.4", "zod-to-json-schema": "^3.25.2" }, "license": "Apache-2.0", "description": "Shared memory engine for AI agents — learn, recall, forget, feedb… Evidence: `packages/core/package.json`
- **Package** (package_manifest): { "name": "@plur-ai/mcp", "mcpName": "io.github.plur-ai/plur", "version": "0.14.0", "type": "module", "bin": { "plur-mcp": "dist/index.js" }, "main": "dist/index.js", "files": "dist", "packs" , "scripts": { "build": "tsup", "test": "vitest run" }, "dependencies": { "@plur-ai/core": "workspace: ", "@modelcontextprotocol/server": "2.0.0-beta.4", "@modelcontextprotocol/client": "2.0.0-beta.4", "@modelcontextprotocol/core": "2.0.0-beta.4", "zod": "^3.23.0" }, "devDependencies": { "@types/node": "^25.5.0" }, "license": "Apache-2.0", "description": "PLUR MCP server — persistent memory tools for any AI agent", "keywords": "ai", "memory", "mcp", "model-context-protocol", "agent", "claude", "cursor"… Evidence: `packages/mcp/package.json`
- **PLUR Memory** (skill_instruction): Most memory plugins send your context to a cloud server. PLUR doesn't. Evidence: `packages/claw/clawhub/SKILL.md`
- **Note: every engram in this pack carries pinned: true at the engram** (skill_instruction): Your agent has memory. These habits make it actually useful. Evidence: `packages/mcp/packs/effective-memory/SKILL.md`
- **PLUR Memory** (skill_instruction): Persistent memory for AI agents. Corrections, preferences, and patterns are stored as engrams that strengthen with use and decay when irrelevant. The system gets smarter the longer you use it. Evidence: `skills/plur-memory/SKILL.md`
- **PLUR Session End** (skill_instruction): Run this at the end of a conversation to extract what is worth remembering. Evidence: `skills/plur-session-end/SKILL.md`
- **Plugin** (structured_config): { "name": "plur", "version": "0.9.13", "description": "Persistent, local-first shared memory for AI agents. Corrections, preferences, and conventions persist across sessions, tools, and machines — stored as plain YAML on your disk, with zero-cost local search.", "author": { "name": "PLUR", "url": "https://plur.ai" }, "homepage": "https://plur.ai", "repository": "https://github.com/plur-ai/plur", "license": "Apache-2.0", "keywords": "memory", "agent-memory", "mcp", "local-first", "context-engine" } Evidence: `.claude-plugin/plugin.json`
- **License** (source_file): Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ Evidence: `LICENSE`
- **Changelog** (documentation): Performance + expansion — lean MCP surface by default, new LangChain adapter, cross-platform path fix, dependency security hardening, and four core stability fixes. Evidence: `CHANGELOG.md`
- **PLUR Roadmap** (documentation): PLUR is persistent, local-first memory for AI agents. This roadmap captures the major workstreams currently active. For day-to-day status, see open issues https://github.com/plur-ai/plur/issues . Evidence: `ROADMAP.md`
- **Known Issues** (documentation): This file tracks known, accepted issues for the current release line, with machine-checkable sentinels that gate release.sh and CI. Evidence: `docs/KNOWN_ISSUES.md`
- **Opt-in Engagement Telemetry — Design & Implementation** (documentation): Opt-in Engagement Telemetry — Design & Implementation Evidence: `docs/telemetry-design.md`
- **Installing PLUR for Cline and other agents** (documentation): Installing PLUR for Cline and other agents Evidence: `llms-install.md`
- **ADR-0002: Derived-state provenance — history JSONL is source of truth for observational events** (documentation): ADR-0002: Derived-state provenance — history JSONL is source of truth for observational events Evidence: `docs/adr/ADR-0002-derived-state-provenance.md`
- **PLUR v0.9.x benchmark baseline — 2026-07-08** (documentation): PLUR v0.9.x benchmark baseline — 2026-07-08 Evidence: `docs/benchmarks/2026-07-08-longmemeval-s.md`
- **Local-first AI memory: feature comparison** (documentation): Local-first AI memory: feature comparison Evidence: `docs/benchmarks/feature-comparison.md`
- **Local-first AI memory: Phase 2 benchmark methodology** (documentation): Local-first AI memory: Phase 2 benchmark methodology Evidence: `docs/benchmarks/phase2-methodology.md`
- **PLUR v0.2.1 on LongMemEval — reference measurement** (documentation): PLUR v0.2.1 on LongMemEval — reference measurement Evidence: `docs/reports/longmemeval-v0.2.1.md`
- **Runbook: Human-bridge credentials NPM TOKEN + ClawHub OAuth** (documentation): Runbook: Human-bridge credentials NPM TOKEN + ClawHub OAuth Evidence: `docs/runbooks/credentials.md`
- **Runbook: Consolidating one engram store into another** (documentation): Runbook: Consolidating one engram store into another Evidence: `docs/runbooks/store-consolidation.md`
- **Architecture** (documentation): @plur-ai/claw is the OpenClaw plugin that gives every agent in the gateway persistent memory — automatically, with no per-agent code changes. It's a ContextEngine plugin: OpenClaw calls into it on every session lifecycle event. Evidence: `packages/claw/ARCHITECTURE.md`
- **Architecture** (documentation): @plur-ai/core is the engram engine — the in-process library that stores, searches, decays, and injects engrams. No network, no auth, no tools — just the data model and the algorithms. Evidence: `packages/core/ARCHITECTURE.md`
- **Architecture** (documentation): @plur-ai/mcp exposes @plur-ai/core as a Model Context Protocol server. It's a thin adapter — almost all logic lives in core. This package owns: the MCP transport, the tool-definition surface, the init and setup CLI, and the auto-update / config bootstrap. Evidence: `packages/mcp/ARCHITECTURE.md`
- **PLUR test pyramid** (documentation): This document records the test architecture for plur-ai/plur, closing the documentation gap left by issue 92 CI integration for remote store test levels . The original 92 spec assumed Docker-based integration tests; we shipped a simpler-and-stronger approach using in-process stub servers. Evidence: `docs/test-pyramid.md`
- **Index** (source_file): import { PlurContextEngine, type PlurContextEngineOptions } from './context-engine.js' import { ensureSystemPrompt, PLUR SYSTEM SECTION } from './system-prompt.js' import { checkForUpdate, CapabilityCanary } from '@plur-ai/core' import { recordEvent } from './telemetry-counters.js' import { flushIfNeeded, registerFlushOnExit } from './telemetry-flush.js' ⋮---- function maybeFlushAfter rolledOver: boolean : void ⋮---- function getEngine path: string : PlurContextEngine ⋮---- register api: any Evidence: `packages/claw/src/index.ts`
- **Index** (source_file): import { shouldOutputJson, outputJson, exit } from './output.js' import { parseGlobalFlags, createPlur } from './plur.js' Evidence: `packages/cli/src/index.ts`
- **Plur** (source_file): import { Plur } from '@plur-ai/core' import type { OutputOptions } from './output.js' ⋮---- export interface GlobalFlags extends OutputOptions { path?: string fast?: boolean } ⋮---- export function parseGlobalFlags argv: string : ⋮---- export function createPlur flags: GlobalFlags : Plur Evidence: `packages/cli/src/plur.ts`
- **Decay** (source_file): import type { Engram } from './schemas/engram.js' ⋮---- export function decayedStrength retrievalStrength: number, daysSinceAccess: number, lambda: number = DECAY RATE, : number ⋮---- export function daysSince lastAccessed: string, now?: Date : number ⋮---- export function shouldInject engram: { retrieval strength: number; scope: string; last accessed?: string }, context: { task?: string; scope?: string }, threshold: number = 0.15, : boolean ⋮---- // Scope-matched engrams ALWAYS inject ignore decay ⋮---- export function reactivate currentStrength: number : number ⋮---- export function decayedCoAccessStrength strength: number, daysSinceUpdate: number, lambda: number = 0.01, : number ⋮---- ex… Evidence: `packages/core/src/decay.ts`
- **Bge Small** (source_file): import { makeTransformersAdapter } from './transformers-base.js' import type { EmbedderAdapter } from './types.js' ⋮---- export function makeBgeSmallAdapter : EmbedderAdapter Evidence: `packages/core/src/embedders/bge-small.ts`
- **Index** (source_file): import { logger } from '../logger.js' import { makeMiniLMAdapter } from './minilm.js' import { makeBgeSmallAdapter } from './bge-small.js' import { makeBgeBaseAdapter } from './bge-base.js' import { makeEmbeddingGemmaAdapter } from './embedding-gemma.js' import { makeOpenAI3LargeAdapter } from './openai.js' import type { EmbedderAdapter } from './types.js' ⋮---- export type EmbedderName = typeof EMBEDDER NAMES number ⋮---- export function resetEmbedderCache : void ⋮---- export function getEmbedder name: EmbedderName : EmbedderAdapter ⋮---- function build name: EmbedderName : EmbedderAdapter ⋮---- export function resolveEmbedderName env: NodeJS.ProcessEnv = process.env : EmbedderName ⋮---- e… Evidence: `packages/core/src/embedders/index.ts`
- **Embeddings** (source_file): import type { Engram } from './schemas/engram.js' import type { EmbedRole } from './embedders/types.js' import { engramSearchText } from './fts.js' import { existsSync, readFileSync, mkdirSync } from 'fs' import { join, dirname } from 'path' import { createHash } from 'crypto' import { atomicWrite } from './sync.js' import { logger } from './logger.js' ⋮---- export function readDisabledFromEnv env: Record : string null ⋮---- export interface EmbedderStatus { available: boolean loaded: boolean lastError: string null disabled: boolean disabledReason: string null } ⋮---- export function embedderStatus : EmbedderStatus ⋮---- export function setEmbeddingsEnabled enabled: boolean, reason?: string… Evidence: `packages/core/src/embeddings.ts`
- **Engrams** (source_file): import { EngramSchemaPassthrough, type Engram } from './schemas/engram.js' import { PackManifestSchema, type PackManifest } from './schemas/pack.js' import { logger } from './logger.js' import { atomicWrite } from './sync.js' ⋮---- export function loadEngrams filePath: string : Engram ⋮---- export function saveEngrams filePath: string, engrams: Engram : void ⋮---- export interface LoadedPack { manifest: PackManifest engrams: Engram } ⋮---- function parseSkillMdFrontmatter filePath: string : Record ⋮---- export function loadPack packDir: string : LoadedPack ⋮---- export function loadAllPacks packsDir: string : LoadedPack ⋮---- export function storePrefix scope: string : string ⋮---- export f… Evidence: `packages/core/src/engrams.ts`
- **Episodes** (source_file): import { existsSync, readFileSync } from 'fs' import yaml from 'js-yaml' import type { Episode } from './schemas/episode.js' import type { CaptureContext, TimelineQuery } from './types.js' import { atomicWrite } from './sync.js' ⋮---- function generateEpisodeId : string ⋮---- export function captureEpisode path: string, summary: string, context?: CaptureContext : Episode ⋮---- export function queryTimeline path: string, query?: TimelineQuery : Episode ⋮---- function loadEpisodes path: string : Episode Evidence: `packages/core/src/episodes.ts`
- **Fts** (source_file): import type { Engram } from './schemas/engram.js' ⋮---- export function ftsTokenize text: string : string ⋮---- export function engramSearchText engram: Engram : string ⋮---- export function computeIdf engrams: Engram , queryTokens: string : Map ⋮---- export function ftsScore engram: Engram, queryTokens: string , idfWeights?: Map , avgDocLength?: number : number ⋮---- export function searchEngrams engrams: Engram , query: string, limit = 20 : Engram Evidence: `packages/core/src/fts.ts`
- **Hybrid Search** (source_file): import type { Engram } from './schemas/engram.js' import { searchEngrams } from './fts.js' import { embeddingSearch, embedderStatus } from './embeddings.js' import { rewriteLexicalQuery, isQueryRewriteDisabled } from './intent/rewrite.js' import { logger } from './logger.js' import type { RerankerAdapter } from './rerankers/types.js' import { isRerankerOff, recordRerankerEngaged, recordRerankerFailure, hfCacheDirName } from './rerankers/index.js' ⋮---- export interface HybridSearchResult { engrams: Engram mode: 'hybrid' 'hybrid-degraded' 'bm25-only' embedderError: string null topScore: number null reranked?: number } ⋮---- export interface RerankOptions { reranker?: RerankerAdapter topK?: n… Evidence: `packages/core/src/hybrid-search.ts`
- **Index** (source_file): import { existsSync, readFileSync } from 'fs' import { basename } from 'path' import type { Plur } from '../index.js' import type { FieldMapping, ImportInput, ImportSource, MigrationReport } from './types.js' import { parseGenericContent } from './generic.js' import { parseMem0Content } from './mem0.js' import { parseGpEngramDb } from './gp-engram.js' import { runImport } from './engine.js' import { zepSource, lettaSource } from './stubs.js' ⋮---- function readText input: ImportInput : string ⋮---- export function listImportSources : ImportSource ⋮---- export function getImportSource name: string : ImportSource ⋮---- export interface ImportFromOptions { from: string path: string mapping?: F… Evidence: `packages/core/src/importers/index.ts`
- **Index** (source_file): import { tmpdir } from 'os' import { join, dirname, basename } from 'path' import yaml from 'js-yaml' import { detectPlurStorage, type PlurPaths } from './storage.js' import { IndexedStorage } from './storage-indexed.js' import { PGLiteAdapter } from './storage-pglite.js' import { loadConfig } from './config.js' import { loadEngrams, saveEngrams, generateEngramId, loadAllPacks, storePrefix } from './engrams.js' import { logger } from './logger.js' import { searchEngrams } from './fts.js' import { selectAndSpread, scoreEngramsPublic, formatWithLayer, assignLayer } from './inject.js' import { reactivate } from './decay.js' import { captureEpisode, queryTimeline } from './episodes.js' import {… Evidence: `packages/core/src/index.ts`
- **Index** (source_file): import { logger } from '../logger.js' import { makeBgeRerankerV2M3Adapter } from './bge-reranker-v2-m3.js' import { makeMsMarcoMiniLmL6Adapter } from './ms-marco-minilm-l6.js' import type { RerankerAdapter } from './types.js' ⋮---- export type RerankerName = typeof RERANKER NAMES number ⋮---- export function resetRerankerCache : void ⋮---- export function setCachedReranker name: RerankerName, adapter: RerankerAdapter : void ⋮---- export type RerankerFailureKind = 'corrupt-cache' 'unavailable' ⋮---- export interface RerankerRuntimeStatus { engaged count: number failure count: number lastError: string null lastErrorKind: RerankerFailureKind null lastFailedReranker: string null } ⋮---- const c… Evidence: `packages/core/src/rerankers/index.ts`
- **Scope Routing** (source_file): import type { ScopeMetadata } from './schemas/scope-metadata.js' ⋮---- export interface ScopeSignals { statement: string domain?: string tags?: string } ⋮---- export interface ScopeCandidate { scope: string confidence: number reason: string domainMatch: boolean coverContainsDomain: boolean coverSpecificity: number } ⋮---- function tokenizeStatement text: string : string ⋮---- function normalizeCover cover: string : string ⋮---- / Does value sit under prefix as a dotted namespace? plur.core.security is under plur and plur.core ; plurple is NOT under plur segment boundary, not substring . Exact equality counts. / function isNamespacePrefix prefix: string, value: string : boolean ⋮---- / Squas… Evidence: `packages/core/src/scope-routing.ts`
- **Scope Util** (source_file): export function isSharedScope scope: string : boolean ⋮---- export function isPersonalScope scope: string : boolean ⋮---- export function isScopeWithin scope: string, queryScope: string : boolean Evidence: `packages/core/src/scope-util.ts`
- **Init** (source_file): logger = logging.getLogger "plur hermes" ⋮---- version = pkg version "plur-hermes" ⋮---- version = "0.0.0+unknown" ⋮---- session state: dict str, dict = {} PRUNE AGE SECONDS = 3600 FEEDBACK MIN CONFIDENCE = 0.6 CORRECTION MARKERS = "actually,", "actually ", "no,", "wrong,", "incorrect,", ⋮---- def get trigrams text: str - set str ⋮---- words = text.lower .split ⋮---- def detect injection signal engram text: str, response: str - tuple str None, float ⋮---- response lower = response.lower engram lower = engram text.lower .strip ⋮---- engram tris = get trigrams engram lower ⋮---- response tris = get trigrams response lower overlap = len engram tris & response tris / len engram tris ⋮---- engra… Evidence: `packages/hermes/plur_hermes/__init__.py`
- The remaining 16 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: `docs/adr/README.md`, `.claude-plugin/README.md`, `.openclaw-plugin/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: `docs/adr/README.md`, `.claude-plugin/README.md`, `.openclaw-plugin/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, Install & Quick Start**: importance `high`
  - source_paths: README.md, CLAUDE.md, CONTRIBUTING.md, ROADMAP.md, CHANGELOG.md
- **Core Engine: Engrams, Episodes, Storage & Scopes**: importance `high`
  - source_paths: packages/core/src/index.ts, packages/core/src/engrams.ts, packages/core/src/episodes.ts, packages/core/src/decay.ts, packages/core/src/scope-routing.ts
- **Search, Recall, Injection & Recall Quality**: importance `high`
  - source_paths: packages/core/src/fts.ts, packages/core/src/embeddings.ts, packages/core/src/embedders/index.ts, packages/core/src/embedders/bge-small.ts, packages/core/src/hybrid-search.ts
- **Integrations: MCP, CLI, OpenClaw, Hermes, Python & LangChain**: importance `high`
  - source_paths: packages/mcp/src/index.ts, packages/mcp/src/tools.ts, packages/mcp/src/server.ts, packages/mcp/ARCHITECTURE.md, packages/cli/src/index.ts

## Repo Inspection Evidence

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `c8208e8505de1e2b2e64d9f0c8a39ca0afac7bce`
- inspected_files: `README.md`, `package.json`, `pnpm-lock.yaml`, `docs/KNOWN_ISSUES.md`, `docs/adr/ADR-0002-derived-state-provenance.md`, `docs/adr/README.md`, `docs/benchmarks/2026-07-08-longmemeval-s.md`, `docs/benchmarks/feature-comparison.md`, `docs/benchmarks/phase2-methodology.md`, `docs/reports/longmemeval-v0.2.1.md`, `docs/runbooks/credentials.md`, `docs/runbooks/store-consolidation.md`, `docs/telemetry-design.md`, `docs/test-pyramid.md`, `examples/01-basic-store-recall.ts`, `examples/02-mcp-integration.ts`, `examples/03-scope-filtering.ts`, `examples/README.md`, `examples/package.json`, `packages/claw/ARCHITECTURE.md`

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: Security or permission risk requires verification

- Trigger: Developers should check this security_permissions risk before relying on the project: Proactively offer to register authorized-but-unregistered scopes (per-scope, remembered opt-out)
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: Proactively offer to register authorized-but-unregistered scopes (per-scope, remembered opt-out). Context: Source discussion did not expose a precise runtime context.
- Why it matters: Developers may expose sensitive permissions or credentials: Proactively offer to register authorized-but-unregistered scopes (per-scope, remembered opt-out)
- Evidence: failure_mode_cluster:github_issue | https://github.com/plur-ai/plur/issues/647
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

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

- Trigger: Developers should check this security_permissions risk before relying on the project: Scope-metadata edits not reflected in running clients (cached until next /me pull) — freshness signal?
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: Scope-metadata edits not reflected in running clients (cached until next /me pull) — freshness signal?. Context: Observed during version upgrade or migration.
- Why it matters: Developers may expose sensitive permissions or credentials: Scope-metadata edits not reflected in running clients (cached until next /me pull) — freshness signal?
- Evidence: failure_mode_cluster:github_issue | https://github.com/plur-ai/plur/issues/648
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.

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

- Trigger: Developers should check this security_permissions risk before relying on the project: plur sync: scope-aware push — don't send private-visibility engrams to shared remotes
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: plur sync: scope-aware push — don't send private-visibility engrams to shared remotes. Context: Source discussion did not expose a precise runtime context.
- Why it matters: Developers may expose sensitive permissions or credentials: plur sync: scope-aware push — don't send private-visibility engrams to shared remotes
- Evidence: failure_mode_cluster:github_issue | https://github.com/plur-ai/plur/issues/640
- 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: Upgrade MCP SDK v2 from 2.0.0-beta.4 to stable GA when released
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: Upgrade MCP SDK v2 from 2.0.0-beta.4 to stable GA when released. Context: Observed when using node
- Why it matters: Developers may fail before the first successful local run: Upgrade MCP SDK v2 from 2.0.0-beta.4 to stable GA when released
- Evidence: failure_mode_cluster:github_issue | https://github.com/plur-ai/plur/issues/645
- 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: compat: MCP SDK v2 migration — import split + Zod v3→v4 (deadline 2026-07-28)
- Host AI rule: Before packaging this project, run the relevant install/config/quickstart check for: compat: MCP SDK v2 migration — import split + Zod v3→v4 (deadline 2026-07-28). Context: Observed when using node
- Why it matters: Developers may fail before the first successful local run: compat: MCP SDK v2 migration — import split + Zod v3→v4 (deadline 2026-07-28)
- Evidence: failure_mode_cluster:github_issue | https://github.com/plur-ai/plur/issues/637
- Hard boundary: Do not present this pitfall as solved, verified, or ignorable unless later evidence explicitly closes it.