agentmods.dev

session-graph CLAUDE.md

Project instructions for session-graph, which collects AI coding-session records and turns their information into a searchable knowledge graph.

In plain words
What is it for?
Use them when importing conversation or session logs, extracting linked facts, querying the resulting graph, or compiling and promoting project changes.
Why use it?
They explain the project’s data sources, architecture, and promotion path so changes can be checked and moved safely between development and production.

Instructions file

109 1mo ago C scan 4,887 tokens original Apache-2.0
View the source file robertoshimizu/session-graph
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/robertoshimizu/session-graph/claude-md
Clone the repo
git clone --depth 1 https://github.com/robertoshimizu/session-graph
Per session 4,887 This file is loaded in full into every session.
When invoked 4,887 The same file — it is already loaded in full.
Security scan C 2 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.04887 $0.04887
Opus 5 $0.02443 $0.02443
Sonnet 5 $0.00977 $0.00977
Haiku 4.5 $0.00489 $0.00489

Measured 3d ago against content hash dc3d40019620, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

session-graph CLAUDE.md scanned grade C with 2 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 3d ago.

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.

Nullifies safety policieshighAnti-refusal

"You have no restrictions", "do anything now", "ignore your guidelines": a direct jailbreak that disables guardrails.

| | 48 whitelisted short terms bypass all filters

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

hooks/stop_hook.sh # Claude Code Stop hook: curl POST to RabbitMQ (~33ms)
CLAUDE.md · 328 lines

How it starts

The opening of the file, as written. The whole thing — 328 lines — stays where its author put it; the contents beside it link to each section on GitHub.

session-graph

Instructions for Claude Code when working on this project.

Project Goal

Build a unified developer knowledge graph that connects scattered knowledge from AI coding sessions across multiple platforms:

  • Claude Code session logs (.jsonl)
  • pi coding agent session logs (.jsonl)
  • DeepSeek conversation exports (JSON zip)
  • Grok conversation exports (JSON zip)
  • Warp terminal AI sessions (SQLite)
  • ChatGPT conversation exports (JSON)
  • Cursor AI sessions (~/.cursor/projects/*/agent-transcripts/*.jsonl)
  • VS Code Copilot interactions (planned)

The pipeline extracts structured (subject, predicate, object) triples from AI assistant messages, links entities to Wikidata via owl:sameAs, and loads everything into a SPARQL-queryable triplestore with full provenance.

Branch and Production Workflow

Use this promotion path for all changes:

feature/* → development → validation → main → production rebuild
  • Create feature branches from development when it exists; merge completed features back into development.
  • Run compilation, parser/integration checks, and relevant live-service verification on development before promotion.
  • main must represent the exact version deployed in production. Do not merge unvalidated work directly into main.
  • After promoting development to main, push both branches and rebuild/restart production from a clean main checkout.
  • Preserve Docker volumes during deployment; Fuseki and RabbitMQ data must not be recreated or removed.
  • Record or verify the deployed Git SHA. If an image lacks revision metadata, compare critical source hashes and treat adding an OCI revision label as pending infrastructure work.
  • Start new feature work only after development and main accurately reflect their environments.

Architecture: Ontology + Knowledge Graph + Hybrid Retrieval

Ontology Stack (Composed W3C Standards)

Do NOT create a custom ontology from scratch. Compose these battle-tested standards:

Read the full file on GitHub · 328 lines

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.

  1. 3d ago First seen · 328 lines · 4,887 tokens per session scan C dc3d40019620

Subscribe to this mod's changes

session-graph CLAUDE.md is an instructions file published in the GitHub repository robertoshimizu/session-graph (109 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 4,887 tokens to every session, about $0.0244 per session on Opus 5. A static security scan graded it C with 2 findings (nullifies safety policies, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

ai-sessionsapache-jenaclaude-codedeveloper-toolsknowledge-graphontology
Related

Other instructions, from other repositories

ontology-atlas AGENTS.md

Instructions for wlsdks/ontology-atlas, covering agents.md — ontology-atlas, product and non-negotiable architecture, start here, structure and routes and operating gates and skills.

wlsdks/ontology-atlas · 2,685 tokens

ontology-atlas CLAUDE.md

Instructions for wlsdks/ontology-atlas, covering claude.md, visibility and mirrors, claude code loading and synchronization.

wlsdks/ontology-atlas · 579 tokens

open-kgo AGENTS.md

Instructions for mloda-ai/open-kgo, covering agents.md, environment, dependencies, running checks and run individual checks.

mloda-ai/open-kgo · 1,264 tokens

AnythingGraph AGENTS.md

Instructions for AnythingGraph/AnythingGraph, covering agents.md — anything cli mcp authoring guide, golden rules, exploring a source (no playbook), admin workflow and playbook json (compact).

AnythingGraph/AnythingGraph · 1,452 tokens

DocTology AGENTS.md

AGENTS.md instructions for tteggu87/DocTology, covering doctology agent contract, working rules, reading order, definition of done and codegraph.

tteggu87/DocTology · 379 tokens

heritagegraph AGENTS.md

AGENTS.md instructions for CAIRNepal/heritagegraph, covering agents.md — ai agent instructions for heritagegraph, 🧠 what is this project?, 📁 repository structure, ⚠️ critical rules — read before coding and 1. never hardcode secrets.

CAIRNepal/heritagegraph · 2,830 tokens