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/evaneos/agent-callable/agents-mdgit clone --depth 1 https://github.com/Evaneos/agent-callableWhat 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 | $0.01122 | $0.01122 |
| Opus 5 | $0.00561 | $0.00561 |
| Sonnet 5 | $0.00224 | $0.00224 |
| Haiku 4.5 | $0.00112 | $0.00112 |
Grade A, and why
agent-callable 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 2d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dev directives for agent-callable
CLAUDE.md is a symlink to AGENTS.md.
This file contains directives for developing and maintaining agent-callable.
For a runtime usage example in another project, see SAMPLE_CLAUDE.md.
Tool purpose (context for design decisions)
agent-callableis a deny-by-default command filter designed to be pre-approvable in an LLM agent's allowlist.- Goal: reduce the surface of accidental side effects from an agent (push/PR/apply/destroy/etc.).
- This is a best-effort filter, not a security boundary. It offers no guarantee against intentional or creative bypass.
Policy principles (to follow in code)
- Deny-by-default: any new subcommand/flag starts blocked until explicitly classified as "OK".
- Prefer false positives (blocking too much) over false negatives (letting dangerous things through).
- No shell: never evaluate shell; execute the binary directly (reduces injection risk).
- Non-interactive:
- neutralize
stdinif TTY when relevant, - disable prompts (e.g.
GIT_TERMINAL_PROMPT=0,GH_PROMPT_DISABLED=1), - avoid/neutralize pagers (e.g.
PAGER=catwhen useful).
- neutralize
- LLM-friendly messages: on block, return a short and actionable error (without requesting input).
- Secrets: block modes/flags that exfiltrate secrets (e.g.
kubectl get secrets -o yaml/json,pulumi --show-secrets). - Persistent changes: block commands that durably modify config (e.g.
helm repo add,gcloud config set) unless explicitly decided. - Remote effects: block remote/destructive operations (push, PR create/merge, apply/destroy, etc.).
- Local "cache/artifact" writes: usually OK if useful for investigation (e.g.
git fetch,docker pull,gh repo clone), as long as it doesn't durably reconfigure the tool.
Before writing or reviewing a tool filter
- Read the tool's official docs/
--helpoutput to know all flags, subcommands, and aliases. Don't rely on memory — CLIs have flags you don't expect. - Keep the security model in mind: README.md § "What gets filtered" defines the three categories (remote effect / persistent config / local cache). Every decision must map to one.
- Don't forget universally harmless patterns:
--version,--help, noop/dry-run flags. These are safe for any tool.
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.
- 2d ago First seen · 70 lines · 1,122 tokens per session scan A 83fafa199c20
agent-callable AGENTS.md is an instructions file published in the GitHub repository Evaneos/agent-callable (5 stars, last pushed 26d ago), licensed MIT. It adds 1,122 tokens to every session, about $0.0056 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
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
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
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, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
langchain 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.