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.
npx agentmods add instructions/ajbarea/ariadne/agents-mdgit clone --depth 1 https://github.com/ajbarea/ariadneWhat 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 | $0.00617 | $0.00617 |
| Opus 5 | $0.00309 | $0.00309 |
| Sonnet 5 | $0.00123 | $0.00123 |
| Haiku 4.5 | $0.00062 | $0.00062 |
Grade A, and why
ariadne AGENTS.md 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 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.
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.
How it starts
The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Ariadne
Repository conventions for agentic coding tools (Claude Code, Cursor, Codex,
Aider). Humans: start with README.md, then ROADMAP.md.
What this project is
A sensemaking harness for nonatomic entities, built on the Claude Agent SDK
as an orchestration layer over heterogeneous, multimodal data. Architecture is
research-driven; see docs/research/. Do not introduce
runtime dependencies or pin connectors/libraries until the decision is recorded
in ROADMAP.md with a # research(YYYY-MM): provenance note.
Build & run
uv is the canonical toolchain; make targets wrap it. make help lists all.
| Goal | Command |
|---|---|
| Install deps | make setup |
| Auto-fix (format + lint) | make fix |
| Lint (CI-equivalent) | make lint |
| Unit tests (fast) | make test-unit |
| Full test suite | make test |
| Pre-push gate | make validate |
| Run the CLI | uv run ariadne |
| Archived/auditable run | ./scripts/dev-runner.sh <target> |
Conventions
- Python ≥ 3.12, src-layout package under
src/ariadne/. - Formatting/linting: ruff (format + check). Types: ty. Config in
pyproject.toml. - Fix-first gates:
make fixbeforemake lintso lint measures intent. - Tests: pytest,
asyncio_mode = auto. Mark integration tests@pytest.mark.integrationand heavy ones@pytest.mark.slowso the fast lane stays hermetic. - No secrets or source intelligence data in the repo.
data/,models/,.env, andlogs/are gitignored. - Provenance matters. This is an analytic tool: every fact the harness surfaces should be traceable to its source tool. Favor designs that preserve citations end-to-end.
Where things go
- Agent tools (custom/in-process connectors & processors) →
tools/. - Agent Skills (packaged analytic procedures) →
skills/<name>/SKILL.md. - MCP servers (graph / SQL / vector connectors) →
mcp_servers/. - Research backing decisions →
docs/research/.
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.
- 3d ago First seen · 54 lines · 617 tokens per session scan A ffaa45f19bdf
ariadne AGENTS.md is an instructions file published in the GitHub repository ajbarea/ariadne (2 stars, last pushed 9d ago), licensed MIT. It adds 617 tokens to every session, about $0.0031 per session on Opus 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.
Other instructions, from other repositories
LightRAG AGENTS.md
AGENTS.md instructions for HKUDS/LightRAG, covering repository guidelines, project overview, project structure, module layout (lightrag/) and core architecture.
edgequake AGENTS.md
AGENTS.md instructions for raphaelmansuy/edgequake, covering repository guidelines, project structure & module organization, build, test, and development commands, quick start with make and start with ollama (default).
jurisd AGENTS.md
Instructions for russellbrenner/jurisd, covering agent instructions for jurisd, purpose, current foundation architecture, command contract rule and mcp surface rule.
hugegraph-ai AGENTS.md
Instructions for apache/hugegraph-ai, covering agents.md, stack & modules, testing expectations, code search anchors and build & test.
LightRAG CLAUDE.md
Claude Code instructions for HKUDS/LightRAG, a project described as:
myrmion-graphrag CLAUDE.md
Claude Code instructions for e2its/myrmion-graphrag, covering claude.md, qué es esto, herramientas del servidor myrmion-graphrag (documentos), herramientas del servidor myrmion-codebase (código) and mantenimiento obligatorio del codebaseinventory.