Borrowing it
Nothing to install: this file belongs to IgorGanapolsky/ThumbGate. 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/IgorGanapolsky/ThumbGate/main/AGENTS.mdgit clone --depth 1 https://github.com/IgorGanapolsky/ThumbGateWrote 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/igorganapolsky/thumbgate/agents-md)<a href="https://agentmods.dev/instructions/igorganapolsky/thumbgate/agents-md"><img src="https://agentmods.dev/badge/instructions/igorganapolsky/thumbgate/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/igorganapolsky/thumbgate/agents-md"><img src="https://agentmods.dev/badge/instructions/igorganapolsky/thumbgate/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 | $0.05124 | $0.05124 |
| Opus 5.5 | $0.02050 | $0.02050 |
| Sonnet 5.5 | $0.01025 | $0.01025 |
| Haiku 4.5 | $0.00512 | $0.00512 |
Grade A, and why
ThumbGate 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 yesterday.
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 — 321 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ThumbGate — The Infrastructure Firewall
npm package:
thumbgate| Brand: ThumbGate
Session PR Management and System Hygiene
At session start, run bin/agent-loop --health --json (fail closed), then read
CLAUDE.md, AGENTS.md, and GEMINI.md; query the local ThumbGate memory;
inspect every open PR, every open GitHub Issue, remote branch, worktree, and
current main CI run before changing repository state.
In PR-management sessions, reconcile GitHub Issues and Actions plus any
available Linear AI and local Obsidian coordination surfaces. Close or advance
actionable Issues in the same session — do not leave the Issues board untouched
while only draining PRs. If Linear, ~/Documents/AI-Agent-Sync, or the
standalone copilot CLI is unavailable, record that fact and continue with
deterministic gh, git, and repository checks. Copilot advice is never
authoritative.
Fleet Coordination Contract (Mandatory, Every Agent, Every Session)
Multiple agents (Claude Code, Codex, grok, Gemini, Hermes) work this repo and
the ~/Documents/AI-Agent-Sync Obsidian vault concurrently. Before ANY state
mutation — checkout switch, clean/reset, branch or worktree creation, push, PR
open/merge, or vault write — complete the coordination sweep:
bash .claude/scripts/session-bootstrap/fleet-coordination-check.sh
The sweep (read-only) reports: the single-writer checkout lease, active vault
claims touching ThumbGate, vault dirty-writer signals, live herdr panes (when
the gateway is up), Linear issues in flight, and the open-PR census. Skill:
.claude/skills/fleet-coordination/SKILL.md (also wired into
.claude/settings.json SessionStart so Claude Code runs it automatically).
Non-negotiable rules:
- One live session owns a checkout (
node scripts/session-lease.js claim). A live foreign lease means use a separate worktree, not a force-checkout. - Never
git add -Ain a shared checkout. Stage explicit paths only. - Check
Agent-Jobs/running/in the vault before claiming or editing shared project files; treat live claims as owned until stale (>7d or dead PID). - Never write to another agent's state file. Commit only your own vault file via PR; the vault has its own branch protection.
- Do not duplicate an open PR or an open Linear issue. Claim first, then work.
- A handoff is a saved state-and-next file. A new session keeps the old one. Children get disjoint files and do not inherit the parent transcript. Observation is view-only. Do not install omp or open a link that can drive the host. Check with
npx thumbgate omp-coordination-honesty --json --map-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.
- yesterday Changed · -323 lines · -5,741 tokens per session fcc30f18be83
- 2d ago Changed · +4 lines · +173 tokens per session 76885fa6c514
- 4d ago Changed · +99 lines · +2,884 tokens per session 5c6321d93772
- 8d ago Changed · +10 lines · +143 tokens per session b4d9f22d1304
- 11d ago Changed · +31 lines · +594 tokens per session 58dabb34066a
- 15d ago Changed · +13 lines · +234 tokens per session e098168aca19
- 22d ago Changed · +67 lines · +908 tokens per session 96ddfb5ea36b
- 1mo ago Changed · +11 lines · +149 tokens per session 748c13067c5a
- 1mo ago Changed · +11 lines · +108 tokens per session 76f4c3816dbb
- 1mo ago First seen · 398 lines · 5,672 tokens per session scan A 63ed841dd6b8
ThumbGate AGENTS.md is an instructions file published in the GitHub repository IgorGanapolsky/ThumbGate (29 stars, last pushed yesterday), licensed MIT. It adds 5,124 tokens to every session, about $0.0205 per session on Opus 5.5. 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 2026-08-30.
Other instructions, from other repositories
plan-forge version.instructions.md
Version management — Semantic versioning, auto-increment, commit-driven version bumps, release tagging.
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).