What is internal-rag?

A project-memory system that stores and retrieves facts, decisions, lessons, and unverified ideas for coding work. RAG means retrieving relevant stored information when a task begins.

Use it at task start, after failures or milestones, before risky changes, and when recovering unfinished work. It can search memories by type or status and create checkpoints.

It reduces repeated explanations and helps prevent earlier mistakes from being repeated. It also marks uncertain information so it can be checked before use.

How PeterPirog/mcp-light-memory configures Claude Code and Codex

Reuse

Borrowing it

Nothing to install: this file belongs to PeterPirog/mcp-light-memory. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/PeterPirog/mcp-light-memory/main/.agents/skills/internal-rag/SKILL.md

Clone the repo

git clone --depth 1 https://github.com/PeterPirog/mcp-light-memory

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/peterpirog/mcp-light-memory/internal-rag/github.svg)](https://agentmods.dev/skills/peterpirog/mcp-light-memory/internal-rag)

Your own site

<a href="https://agentmods.dev/skills/peterpirog/mcp-light-memory/internal-rag"><img src="https://agentmods.dev/badge/skills/peterpirog/mcp-light-memory/internal-rag/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/peterpirog/mcp-light-memory/internal-rag"><img src="https://agentmods.dev/badge/skills/peterpirog/mcp-light-memory/internal-rag.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 62 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,680 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00062	$0.01680
Opus 5.5	$0.00025	$0.00672
Sonnet 5.5	$0.00012	$0.00336
Haiku 4.5	$0.00006	$0.00168

Measured 2mo ago against content hash e3c169a28ce5, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

internal-rag scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

The scan reads SKILL.md. This mod also ships 11 executable files (irag_atomic.py, irag_distill.py, irag_embeddings.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/internal-rag/SKILL.md · 184 lines

How it starts

The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Internal RAG

Start every substantial task

Windows:

python .agents\skills\internal-rag\mlm.py context --task "<current task>"

Linux/macOS:

python3 .agents/skills/internal-rag/mlm.py context --task "<current task>"

The context packet groups memories by type:

Verified facts (decisions, knowledge, constraints) — trust these as established context.
Lessons & pitfalls (gotchas, failures) — apply to avoid repeating mistakes.
Unverified hypotheses — treat as tentative ideas, not facts. Verify before acting.

If RECOVERY REQUIRED appears, do not make new edits. Inspect Git state, reconstruct unfinished work, checkpoint it, then run guard.

Filtering by type or status

# Only decisions and knowledge (verified facts)
mlm.py search --query "database" --type decision knowledge

# Only active memories (exclude tentative)
mlm.py search --query "cache" --status active

# Only hypotheses (what's still unverified)
mlm.py search --query "performance" --type hypothesis

# Combine in context
mlm.py context --task "fix database pool" --type decision gotcha --status active

Before the first code edit

Create a task-start checkpoint:

python3 .agents/skills/internal-rag/mlm.py checkpoint --reason "task-start" --phase "starting implementation" --next "first concrete implementation step"

On Windows use python instead of python3.

Checkpoint frequently

Checkpoint after each meaningful milestone, important discovery, blocker/failure, or plan change, and before dependency installs, large builds/tests, migrations, broad refactors, compaction, and the final answer.

Guard before finishing

python3 .agents/skills/internal-rag/mlm.py guard

If guard is stale, checkpoint and run guard again. Do not finish until GUARD OK.

Retrieval (selective, never preload all)

python3 .agents/skills/internal-rag/mlm.py search --query "symbols subsystem error" --limit 8
python3 .agents/skills/internal-rag/mlm.py search --query "..." --json   # for tooling

Read the full file on GitHub · 184 lines

Files

What ships with it

11 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2mo ago First seen · 184 lines · 62 tokens per session scan A e3c169a28ce5

Subscribe to this mod's changes

internal-rag is a skill published in the GitHub repository PeterPirog/mcp-light-memory (0 stars, last pushed 2mo ago), licensed MIT. It adds 62 tokens to every session and 1,680 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Data and AI Memory and context RAG and vectors agent-memory coding-agents developer-tools local-first mcp model-context-protocol

internal-rag next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
internal-ragPeterPirog/mcp-light-memory	Skill	0	2mo ago	MIT	A	62	Claude Code, Codex
knowledge-orgbojieli/ai-agent-book	Skill	57k	29d ago	Apache-2.0	A	96	Claude Code, Codex
xerjxerj-org/xerj	Skill	19k	yesterday	Apache-2.0	A	38	Claude Code, Codex
install-openviking-memoryvolcengine/OpenViking	Skill	42k	yesterday	AGPL-3.0	A	191	Claude Code, Codex
kb-memoryfish827-08/rag-kb	Skill	5	2d ago	Apache-2.0	A	112	Claude Code, Codex
setup-open-indexDrDroidLab/open-index	Skill	113	2mo ago	MIT	A	48	Claude Code, Codex
goodmem:helpPAIR-Systems-Inc/goodmem-claude-code-plugin	Skill	9	3d ago	MIT	A	46	Claude Code

mcp-light-memory: Skill for Claude Code