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/tortastudios/python-ai-guardrails-template/claude-mdgit clone --depth 1 https://github.com/tortastudios/python-ai-guardrails-templateWrote 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/tortastudios/python-ai-guardrails-template/claude-md)<a href="https://agentmods.dev/instructions/tortastudios/python-ai-guardrails-template/claude-md"><img src="https://agentmods.dev/badge/instructions/tortastudios/python-ai-guardrails-template/claude-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 | $0.00984 | $0.00984 |
| Opus 5 | $0.00492 | $0.00492 |
| Sonnet 5 | $0.00197 | $0.00197 |
| Haiku 4.5 | $0.00098 | $0.00098 |
Grade A, and why
python-ai-guardrails-template CLAUDE.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 4d 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 — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Code — Quality Contract
How the quality gate works
A PostToolUse hook in .claude/settings.json runs make check automatically
every time you edit or create a .py file. You do not need to run it manually —
the hook handles it.
If the hook fails, you must fix the issue before moving on. Do not proceed to the next task. Do not mark a task complete. Until the check passes, keep fixing.
Fixing failures
When make check fails:
- Read the full output — every line
- Run
make fix FILE=<filepath>first (fixes formatting and safe ruff violations automatically) - Wait for the hook to re-run on the next edit, or manually run
make check FILE=<filepath> - If still failing, fix the remaining issues manually
- Keep editing until the check passes
Never ask the user whether to fix quality issues. Just fix them.
Critical functions
Before editing any function, check CRITICALITY.md.
If the function is listed there:
- Type annotations are mandatory — full input and return types
- Docstring is mandatory — document contract, parameters, return value, and exceptions
- Existing tests must still pass after your change
- New behavior must have new tests
Raise extra scrutiny on high fan-in functions. Many callers means a subtle behavior change can break things in non-obvious ways.
Code style rules
- Prefer early returns over nested conditionals
- Extract helper functions rather than writing inline logic blocks
- Keep functions under 30 lines of actual logic
- Keep files under 300 SLOC — split at natural module boundaries
- Name things for what they are, not what they do (nouns for data, verbs for actions)
- No commented-out code — delete it, git has history
Type annotation rules
When a type annotation exceeds the complexity thresholds enforced by
scripts/check_type_complexity.py (nesting depth > 2 or length > 40 chars),
replace it with a named type using the following decision rules.
When to use each approach
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.
- 4d ago First seen · 131 lines · 984 tokens per session scan A fdaddd8192bc
python-ai-guardrails-template CLAUDE.md is an instructions file published in the GitHub repository tortastudios/python-ai-guardrails-template (11 stars, last pushed 5mo ago), licensed MIT. It adds 984 tokens to every session, about $0.0049 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 2026-08-30.
Other instructions, from other repositories
codeql-development-mcp-server server_ql_languages_tools.instructions.md
Instructions for organizing CodeQL queries and query unit tests for all code languages supported by the CodeQL MCP server tools.
codeql-development-mcp-server server_test_ts.instructions.md
Instructions for MCP server TypeScript test files.
static-analysis AGENTS.md
Instructions for analysis-tools-dev/static-analysis, covering guidelines for ai agents and assistants and 🤖 instructions for the ai.
PhpCodeArcheology CLAUDE.md
Instructions for PhpCodeArcheology/PhpCodeArcheology, covering claude.md — phpcodearcheology, project overview, language, tech stack and project structure.
libthumbor AGENTS.md
AGENTS.md instructions for thumbor/libthumbor, covering agents.md, dev environment setup, running tests, code style and project structure.
Test Generation
Comprehensive Pytest conventions, mocking strategies, and model usage rules for OpenWebUI plugins.