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/giulioder/re-call/claude-mdgit clone --depth 1 https://github.com/GiulioDER/RE-callWrote 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.
[](https://agentmods.dev/instructions/giulioder/re-call/claude-md)<a href="https://agentmods.dev/instructions/giulioder/re-call/claude-md"><img src="https://agentmods.dev/badge/instructions/giulioder/re-call/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.12460 | $0.12460 |
| Opus 5 | $0.06230 | $0.06230 |
| Sonnet 5 | $0.02492 | $0.02492 |
| Haiku 4.5 | $0.01246 | $0.01246 |
Grade B, and why
RE-call CLAUDE.md scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads MCP configurationmediumAgent snooping
mcp.json carries server URLs and auth tokens; reading it lets a mod discover and abuse other integrations.
host's `.env`. `.mcp.json` therefore carries no secret: it sources that `.env` on the far side. Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
URL=$(python -c "import json,urllib.request;d=json.load(urllib.request.urlopen('https://pypi.org/pypi/mypy/2.3.0/json'));print(next(f['url'] for f in d['urls'] if f['filename'].endswith('py3-none-any.whl')))") How it starts
The opening of the file, as written. The whole thing — 753 lines — stays where its author put it; the contents beside it link to each section on GitHub.
recall: working rules
This file is loaded into every session in this repository. It records the rules that have been paid for in lost work, and the two commands a session opens and closes with.
One session, one workspace
Never work in the main checkout, and never in a worktree another session holds.
scripts/session-open.sh enforces both and stops rather than warning.
scripts/session-space.sh new <short-name> # your own worktree, off origin/master, claimed
Why this is a mechanism and not advice: measured 2026-08-15, this repository had twenty
worktrees, five of them holding uncommitted work, and the main checkout itself dirty on a branch two
months behind master. Two sessions in one checkout share an index and a working tree, so one
stages the other's files, a rebase moves HEAD under a session that is mid-edit, and git log looks
fine afterwards because the subjects match. The recorded cost includes a force-push that erased
three already-pushed commits.
- The main checkout is refused outright. It is shared by definition: every worktree resolves its objects through it, and it is the directory every session reaches by habit.
- Each worktree carries a claim naming the session that holds it, written to that worktree's own git directory, so it is never committed and dies with the worktree. A second session is told whose it is and how old the claim is, instead of discovering the collision through a lost edit.
- A claim goes stale after 12 hours with no live process. Liveness is asked of Windows, not of
kill -0, which cannot see a Windows pid from Git Bash and reported a runningclaude.exeas dead. "Cannot tell" counts as alive: a false alive costs arelease --force, a false dead costs somebody's in-flight work. - If
session-space.shis missing,session-open.shstops anyway. A checkout old enough to lack it is exactly the stale shared one, so an absent guard must not read as a passing guard.
scripts/session-space.sh whose prints the current claim. release --force takes over one you are
certain is abandoned.
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.
- today Changed · +89 lines · +1,517 tokens per session c50947ed79a0
- 4d ago First seen · 664 lines · 10,943 tokens per session scan B a8293ad6f54f
RE-call CLAUDE.md is an instructions file published in the GitHub repository GiulioDER/RE-call (3 stars, last pushed 2d ago), licensed Apache-2.0. It adds 12,460 tokens to every session, about $0.0623 per session on Opus 5. A static security scan graded it B with 2 findings (reads mcp configuration, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
OpenNotebookLM AGENTS.md
AGENTS.md instructions for tom1030507/OpenNotebookLM, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
knowledge-base CLAUDE.md
Claude Code instructions for riemannulus/knowledge-base, covering knowledge-base 개발 가이드, 명령어, 아키텍처 (데이터 흐름), 깨뜨리기 쉬운 불변 원칙 and 확장 포인트.
mnemon-memory-mcp CLAUDE.md
Claude Code instructions for nikitacometa/mnemon-memory-mcp, covering mnemon-mcp, commands, tech stack, architecture and database.
fortemi CLAUDE.md
Claude Code instructions for Fortemi/fortemi, covering matric memory, ci/cd, gitea actions workflows, monitoring builds and check recent runs via mcp.
fortemi AGENTS.md
AGENTS.md instructions for Fortemi/fortemi, covering agents.md, framework context, tier 1 / tier 2 / tier 3 loading model, tier 2 capability map and agents.
sie AGENTS.md
AGENTS.md instructions for superlinked/sie, covering agents.md — public sie contributor guide, scope, commands, development boundaries and releases.