Borrowing it
Nothing to install: this file belongs to rjmurillo/ai-agents. 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/rjmurillo/ai-agents/main/.github/instructions/tool-use-hook-bar.instructions.mdgit clone --depth 1 https://github.com/rjmurillo/ai-agentsWrote 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/rjmurillo/ai-agents/tool-use-hook-bar)<a href="https://agentmods.dev/instructions/rjmurillo/ai-agents/tool-use-hook-bar"><img src="https://agentmods.dev/badge/instructions/rjmurillo/ai-agents/tool-use-hook-bar.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 5.1 | $0.02431 | $0.02431 |
| Opus 5 | $0.01215 | $0.01215 |
| Sonnet 5 | $0.00486 | $0.00486 |
| Haiku 4.5 | $0.00243 | $0.00243 |
Grade A, and why
ai-agents tool-use-hook-bar.instructions.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 — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The Tool-Use Hook Bar
Operative rule for ADR-084 rule 6. The ADR carries the decision and its evidence; this file is what to do at the authoring seam.
Scope: every per-call event, vendored or repo-local. That is PreToolUse,
PostToolUse, PermissionRequest, and PostToolUseFailure, matching ADR-084
rule 6. Naming fewer would leave the bar evadable by event choice: a guard
refused on PostToolUse could re-register on PostToolUseFailure, pay the same
spawn, and match every tool.
PostToolUseFailure deserves the most scrutiny, not the least.
build/scripts/generate_dispatcher.py:517 omits it from _MATCHER_EVENTS, so
it cannot be tool-scoped at all: any registration on it fires on every failed
tool call. It is the one per-call event with no narrowing available. Nothing is
registered on it today (ADR-097), so this is a warning about what re-adding one
would cost, not a description of a live hook.
A tool-use hook runs on the caller's hot path on every matching call, for the life of the install. Its cost scales with how hard the agent is working, which the session-boundary events do not.
The cost is a blast radius, not a stopwatch
Lead with this, because the millisecond framing understates the risk by orders of magnitude.
Issue #5013, recorded at .agents/architecture/ADR-085-...asymmetry.md:428-431:
the push_pr_script_identity_guard was registered on a bare Bash matcher, and
combined with the Copilot dispatcher's timed child-process deny (#4706) it
denied 127 unrelated Bash commands over more than 21 minutes before the
owner contained it. That is the real failure mode: a wrong deny on a hot tool
takes out unrelated work. Per-spawn latency is the smaller half.
Three mechanisms make a tool-use matcher wider than it looks. None is visible
from .claude/, which is where hooks are authored.
1. Copilot matches on tool name only. Its matcher targets are toolName for
preToolUse and postToolUse (see
.claude/skills/agent-harness-reference/references/official-hook-contracts.md,
"Matchers"). There is no argument-scoped form. Claude accepts Bash(git push*);
build/scripts/generate_dispatcher.py strips the argument constraint via
_COMMAND_SCOPED_RE, so it degrades to bare Bash and the interpreter starts on
every Bash call, filtering in-process after the spawn.
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 · 180 lines · 2,431 tokens per session scan A 3b951b2e0680
ai-agents tool-use-hook-bar.instructions.md is an instructions file published in the GitHub repository rjmurillo/ai-agents (45 stars, last pushed today), licensed MIT. It adds 2,431 tokens to every session, about $0.0122 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 2026-09-03.
Other instructions, from other repositories
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.
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.
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).
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).
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.
spec-kit AGENTS.md
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.