Borrowing it
Nothing to install: this file belongs to aaif/community-events. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/aaif/community-events/main/AGENTS.mdgit clone --depth 1 https://github.com/aaif/community-eventsWrote 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/aaif/community-events/agents-md)<a href="https://agentmods.dev/instructions/aaif/community-events/agents-md"><img src="https://agentmods.dev/badge/instructions/aaif/community-events/agents-md/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.
<a href="https://agentmods.dev/instructions/aaif/community-events/agents-md"><img src="https://agentmods.dev/badge/instructions/aaif/community-events/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.04065 | $0.04065 |
| Opus 5.5 | $0.01626 | $0.01626 |
| Sonnet 5.5 | $0.00813 | $0.00813 |
| Haiku 4.5 | $0.00407 | $0.00407 |
Grade A, and why
community-events AGENTS.md scanned grade A 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 11d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
`urllib` in `lib/aaif_events/luma.py`. Ops scripts default to read-only — Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
comment, and a third had a bare `subprocess.run` with no retries at all — How it starts
The opening of the file, as written. The whole thing — 264 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to Claude Code (claude.ai/code) when working with code
in this repository. CLAUDE.md is a symlink to it.
README.md is for the person using the toolkit, who is not technical and
runs it from Goose or Claude Desktop. Keep it in plain words: what to ask for,
what is safe, how to set up once. Engine detail, gws setup and the
standalone-zip list belong in MAINTAINING.md. CONTRIBUTING.md covers skill
authoring; DESIGN.md covers how anything this repo renders for a human is
styled. This file is what those don't say.
An installed plugin is not a checkout. Goose installs a git clone of this
repo; Claude Desktop installs a copy without .git, one folder per version.
Both run each script from wherever the session is, with no shell environment. So a SKILL.md command names its script by
<skill-root>, never skills/<name>/... (check_skill_portability.py
enforces this; eleven skills once broke it while working fine from the repo
root). And a secret an engine needs is read from the environment, the
repo-root .env, then the macOS keychain, because the keychain is the one
route that reaches an installed plugin. aaif-check-setup is the operator's
diagnosis for all of it. Nothing an engine writes defaults to the cwd, which
from those apps is the operator's own, often synced, folder: it goes through
aaif_events.datadir (with copies in the lib-free sync.py and
backup.py), and a scratch file such as changes.json goes in mktemp -d.
Every deliberate copy of a lib decision — those two, and check_setup.py's
secret lookup — is pinned to the library by tests/test_pinned_copies.py;
a new copy goes on that list in the same commit.
This repo is public — never commit PII
The skills operate on real organizer and attendee data (Slack directory, intake sheet, chapter CRMs). None of it belongs in git:
- Write audit output, backups, and exports only to paths
.gitignorealready covers (.slack-audit-cache/,**/backups/*,slack-*-audit.*, and the downloaded-into-cwd shapes*.docx/*.xlsx/*.pptx/*.png,changes.json,luma.md,new.md; onlylib/aaif_events/tests/fixtures/andassets/are re-included). Extend.gitignorebefore a script starts emitting a new kind of output. Ignored is not the same as safe — delete these files when the run is done. - Tests and fixtures use synthetic data only —
[email protected],Ada,Boston. Never paste a real row, email, name, Slack ID, handle, phone number or LinkedIn URL into a test, docstring, comment, or commit message. - Never paste a value you just observed in a live run. This is the rule that
actually gets broken, because it does not feel like committing PII — it feels
like showing your work. A real organizer's Gmail address went into a
lib/docstring as evidence for a lookup bug, under the words "Verified live", and two real Slack ids went into a test fixture as "known-good" inputs. Both were written by someone who had read the rule above. Record the shape and the date, never the value: "verified live 2026-09-14: the dotted spelling of a real organizer's address missed where the dotless spelling hit" proves the same thing and names nobody. If a reader would need the real value to trust the claim, the claim belongs in the PR description, not in the repo. - A commit is not undoable here. Rewriting history and force-pushing cleans the branch and nothing else: GitHub keeps serving the old commit at its SHA — verified, not assumed — and only a GitHub Support request garbage-collects it. Treat every identifier that reaches a commit as published permanently, and act accordingly (tell the person, rather than assuming a rewrite covered it).
scripts/check_no_real_pii.pyenforces the decidable half of this on every commit and in CI: a Slack-id shape must be on its allowlist, and an address must be at a synthetic domain (or carry a stand-in local part at one of the domains this estate really uses). It cannot tell a real name from an invented one, so it is a floor, not a guarantee — and adding to its allowlist to make a check pass is the bug, not the fix, unless the value is genuinely invented.- Stage explicitly. A single
git add -Apublishes irreversibly. - Never paste a token into a command the agent runs; put it in
.env(gitignored) or the keychain, and neverexport TOKEN=...interactively — shell history and the transcript both keep it. - Secrets are NEVER accepted as command-line parameters (argv is visible in
psand gets echoed in logs/console) — scripts read tokens and API keys from environment variables (or.env/ keychain) only.
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.
- 11d ago Changed · +22 lines · +426 tokens per session d2a4e9d14c04
- 15d ago Changed · +2 lines · +41 tokens per session 9fbc10208188
- 22d ago Changed · +5 lines · +84 tokens per session 467b701244ef
- 1mo ago Changed · +31 lines · +496 tokens per session 6b605cb65715
- 1mo ago Changed · +25 lines · +422 tokens per session 38071426123c
- 1mo ago Changed · +28 lines · +450 tokens per session 7d2ee82b1087
- 1mo ago Changed · +24 lines · +446 tokens per session a6ad9650e065
- 1mo ago First seen · 127 lines · 1,700 tokens per session scan A b071e48a9dc5
community-events AGENTS.md is an instructions file published in the GitHub repository aaif/community-events (12 stars, last pushed 8d ago), licensed MIT. It adds 4,065 tokens to every session, about $0.0163 per session on Opus 5.5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.