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/nikita3005/taintgate/agents-mdgit clone --depth 1 https://github.com/Nikita3005/taintgateWrote 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/nikita3005/taintgate/agents-md)<a href="https://agentmods.dev/instructions/nikita3005/taintgate/agents-md"><img src="https://agentmods.dev/badge/instructions/nikita3005/taintgate/agents-md.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 | $0.00259 | $0.00259 |
| Opus 5 | $0.00130 | $0.00130 |
| Sonnet 5 | $0.00052 | $0.00052 |
| Haiku 4.5 | $0.00026 | $0.00026 |
Grade A, and why
taintgate 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 5d 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.
What it actually says
TaintGate Engineering Guide
TaintGate is a provenance-aware runtime security layer for AI agents.
Core principle
Untrusted text is data, not authority.
Values originating from webpages, email, retrieval systems, MCP servers, external tools, or other untrusted sources must not silently gain authority to execute privileged side effects.
Engineering priorities
- Security correctness
- Simple developer API
- Framework independence
- Testability
- Minimal dependencies
- Explainable security decisions
Python
- Python >= 3.10
- Use type hints
- Prefer standard library functionality
- Avoid unnecessary dependencies
- Keep public APIs small
- Preserve wrapped function behavior
Security
Never silently fail open.
Every security decision must have structured, explainable reasons.
Never log raw secret values.
Treat externally sourced text as potentially hostile.
Tests
Every feature must include tests.
Before completing work run:
pytest ruff check .
Never weaken security behavior merely to make a test pass.
Architecture
Core security logic must remain framework independent.
OpenAI Agents, LangGraph, CrewAI, MCP and other integrations belong in dedicated adapters.
Git
Keep changes focused.
Do not modify unrelated files.
Avoid broad refactors while implementing small features.
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.
- 5d ago First seen · 65 lines · 259 tokens per session scan A d0a843750a57
taintgate AGENTS.md is an instructions file published in the GitHub repository Nikita3005/taintgate (5 stars, last pushed 17d ago), licensed MIT. It adds 259 tokens to every session, about $0.0013 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
ziran CLAUDE.md
Claude Code instructions for taoq-ai/ziran, covering ziran development guidelines, active technologies, project structure, commands and code style.
deepagents AGENTS.md
AGENTS.md instructions for langchain-ai/deepagents, covering global development guidelines for the deep agents monorepo, corridor security analysis, development workflow, suppressing ruff rules and pr conventions.
Doberman-Core AGENTS.md
AGENTS.md instructions for DobermanCore/Doberman-Core, covering claude.md — doberman operating manual, 0. on startup (every session), 1. what this repository is, 2. architecture & extension points and 3. prime directives (non-negotiable).
arcjet-js AGENTS.md
AGENTS.md instructions for arcjet/arcjet-js, covering agent guidance, examples live in arcjet/examples, agent skills and integration work: review before a pr.
dawnai AGENTS.md
AGENTS.md instructions for cacheplane/dawnai, covering agents.md, what this is (and isn't), workspace map, definition of done and conventions.
fullstack-langgraph-nextjs-agent copilot-instructions.md
Copilot instructions for agentailor/fullstack-langgraph-nextjs-agent, covering ai agent instructions, architecture overview, core agent system, data flow and essential development commands.