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/heretek-ai/agent-proof/gemini-mdgit clone --depth 1 https://github.com/Heretek-AI/Agent-ProofWhat 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 | $0.01336 | $0.01336 |
| Opus 5 | $0.00668 | $0.00668 |
| Sonnet 5 | $0.00267 | $0.00267 |
| Haiku 4.5 | $0.00134 | $0.00134 |
Grade B, and why
Agent-Proof GEMINI.md scanned grade B with 1 finding 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.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
| **Installer** | [`src/installer/`](file:///home/john/Projects/Agent-Proof/src/installer/) | Git hook installation and read-only permission lock-in (`chmod 0444`) | How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GEMINI.md — Google Gemini & Antigravity Assistant Guide
This document defines workflows, tool execution guidelines, and architectural standards for Gemini / Antigravity sessions working on the @heretek-ai/agent-proof codebase.
🛠️ Project Ecosystem & Stack
- Package:
@heretek-ai/agent-proof - Repository:
https://github.com/Heretek-AI/Agent-Proof - Registry:
https://registry.npmjs.org/@heretek-ai/agent-proof - Runtime: Node.js
>= 18.0.0 - Build System:
tsup(ESBuild-backed dual CJS/ESM packager) - Test Framework:
vitest - Compiler: TypeScript 5.7+ with strict NodeNext module resolution
📋 Core Development Workflows
1. Build and Typecheck
# Build CJS, ESM, and .d.ts declarations into dist/
npm run build
# Run strict TypeScript compiler verification without emitting
npm run typecheck
2. Testing & Verification
# Run unit & integration test suites (16 suites, 86 tests)
npm test
# Run real-world sandbox lifecycle test
npm run test:real-repo
# Run polyglot GitHub matrix verification across 5 real-world repositories
npm run test:matrix
# Run oneshot 5-issue real-world Drop verification
npm run test:e2e-5-issues
# Run automated real-world E2E test on Heretek-AI/drop
npm run test:e2e-drop
3. Publishing & CI/CD
- GitHub Actions workflow
.github/workflows/publish.ymluses Trusted Publishing (OIDC) withid-token: writeand npm provenance. - Publishing is triggered by creating a GitHub Release, pushing a version tag (
v*.*.*), or manualworkflow_dispatch. - Version resolution queries the NPM registry and automatically increments the patch version to prevent
EPUBLISHCONFLICT.
🧠 Codebase Knowledge & Key Modules
| Module | Location | Purpose |
|---|---|---|
| Types | src/types/index.ts |
Canonical interfaces, LSP & SARIF schemas, diagnostic envelopes |
| Detector | src/detector/stackDetector.ts |
Multi-stack inspection across polyglot repositories |
| Generator | src/generator/configGenerator.ts |
Deterministic config templates (lefthook.yml, .claude/hooks.json, biome.json, ruff.toml, .aislop/config.yml) |
| ByteFence | src/broker/byteFence.ts |
Pre-write broker, preimage verification, and specification/test freezing |
| Sanitizer | src/sanitizer/lspSanitizer.ts |
Agentjacking defense & command injection scrubber |
| Formatter | src/formatter/diagnosticStream.ts |
ANSI stripper, tool output parsers, LSP envelope builder |
| SARIF | src/formatter/sarifStream.ts |
OASIS SARIF v2.1.0 formatter with exact line/col replacement regions |
| Attestation | src/attestation/provenance.ts |
In-toto cryptographic provenance & Ed25519 signing engine |
| LoopBreaker | src/runner/loopBreaker.ts |
Failure loop tripwire preventing token and context window exhaustion |
| Installer | src/installer/ |
Git hook installation and read-only permission lock-in (chmod 0444) |
| Runner | src/runner/gateRunner.ts |
Gate stage runner for post-edit, pre-commit, pre-push |
| CLI | src/cli.ts |
Command dispatcher (init, detect, run, freeze, unfreeze, sanitize, attest, lock, unlock, status) |
| Launcher | bin/agent-proof.js |
Zero-dependency binary resolution with JS fallback |
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 · 82 lines · 1,336 tokens per session scan B b918f64af2b6
Agent-Proof GEMINI.md is an instructions file published in the GitHub repository Heretek-AI/Agent-Proof (1 stars, last pushed 12d ago), licensed MIT. It adds 1,336 tokens to every session, about $0.0067 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
toh-framework CLAUDE.md
Instructions for wasintoh/toh-framework, covering claude.md — toh framework (repo development guide), what this is, everyday commands, verification protocol and single source, transformed per ide.
awesome-copilot-id AGENTS.md
Instructions for GulajavaMinistudio/awesome-copilot-id, covering communication, explanation and documentation, markdown formatting, user communication style and workflow & methodology.
codevira CLAUDE.md
Instructions for sachinshelke/codevira, covering codevira — persistent project memory, when to call which codevira tool, at the start of every session, before modifying any file and before adopting a pattern, library, or naming convention.
agent-skills AGENTS.md
AGENTS.md instructions for racastellanosm/agent-skills, covering agents.md, 1. vision & architectural invariants, 2. repository architecture & layout, 3. harness support & installation paths and 4. scripting & tooling guardrails.
everything-gemini-code GEMINI.md
Instructions for Jamkris/everything-gemini-code, covering project instructions, prompt defense baseline and guidelines.
ZipAI CLAUDE.md
Claude Code instructions for nickdesi/ZipAI, covering claude.md — zipai: ultra-dense token optimizer, rules, 1. zero filler, 2. ambiguity and 3. prompt caching.