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/oaslananka/debug-recorder-mcp/agents-mdgit clone --depth 1 https://github.com/oaslananka/debug-recorder-mcpWrote 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/oaslananka/debug-recorder-mcp/agents-md)<a href="https://agentmods.dev/instructions/oaslananka/debug-recorder-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/oaslananka/debug-recorder-mcp/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 | $0.01292 | $0.01292 |
| Opus 5 | $0.00646 | $0.00646 |
| Sonnet 5 | $0.00258 | $0.00258 |
| Haiku 4.5 | $0.00129 | $0.00129 |
Grade A, and why
debug-recorder-mcp 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 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 — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior software engineer.
WORKING PRINCIPLES:
[ENVIRONMENT — BEFORE STARTING THE TASK]
- Detect the operating system:
- Linux/macOS:
uname -s && echo $SHELL && node --version - Windows:
$env:OS && $PSVersionTable.PSVersion && node --version
- Linux/macOS:
- Confirm required tools are available (
git,node,npm, optionalpython/pip). - Check whether
AGENTS.mdexists. If it does, read it and follow its rules. - Check whether
README.mdexists. If it does, read it.
[PIN LIBRARIES]
Before writing code, research current stable versions and update the approved table.
Approved Dependencies
| Package | Version | Why |
|---|---|---|
@modelcontextprotocol/sdk |
1.29.0 |
Current stable release on the MCP SDK v1.x line as of April 2026. |
better-sqlite3 |
12.8.0 |
Current stable native SQLite release compatible with the supported Node 22 and Node 24 LTS lines. |
fuse.js |
7.3.0 |
Current stable Fuse.js release compatible with the existing fuzzy search API. |
zod |
3.25.76 |
Stable patch release that satisfies the 3.25+ requirement and behaves well with TypeScript inference. |
eslint |
9.39.4 |
Current stable release on the ESLint 9 maintenance line, selected for flat config migration. |
@eslint/js |
9.39.4 |
Kept on the same maintenance line as ESLint 9 flat config. |
@typescript-eslint/eslint-plugin |
8.58.1 |
Current stable plugin release compatible with ESLint 9 and TypeScript 5.9. |
@typescript-eslint/parser |
8.58.1 |
Parser release compatible with ESLint 9 and TypeScript 5.9. |
typescript |
5.9.3 |
Current TypeScript 5.9 patch line for the conservative quality pass. |
@types/node |
22.19.17 |
Stable Node type line for the Node 22 support floor while runtime CI also covers Node 24 LTS. |
globals |
17.4.0 |
Required to define environment globals explicitly in ESLint flat config. |
ts-node |
10.9.2 |
Required helper tool for npm run dev. |
@jest/globals |
30.4.1 |
Current stable explicit Jest globals package used by TypeScript test imports. |
@types/jest |
30.0.0 |
Current stable Jest type package aligned with the Jest 30 test runtime. |
jest |
30.4.2 |
Current stable Jest release used to keep the test runner on the maintained major line. |
jest-junit |
17.0.0 |
Current stable release with patched uuid transitive dependency, required to publish Azure DevOps test results as JUnit XML. |
ts-jest |
29.4.11 |
Current stable transformer release compatible with Jest 30 and TypeScript 5.9. |
fast-check |
4.8.0 |
Current stable property-based testing release for Jest-compatible TypeScript fuzz/regression tests. |
knip |
6.14.2 |
Current stable dead-code and dependency checker compatible with the supported Node 22 and Node 24 LTS lines. |
babel-plugin-istanbul |
8.0.0 |
Current stable coverage instrumentation plugin used through overrides to avoid deprecated transitive glob paths. |
glob |
13.0.6 |
Current stable glob release used through overrides to keep Jest coverage tooling off deprecated glob versions. |
test-exclude |
8.0.0 |
Current stable coverage exclusion helper compatible with babel-plugin-istanbul 8 and glob 13. |
typedoc |
0.28.18 |
API documentation generator compatible with TypeScript 5.9. |
typedoc-plugin-markdown |
4.11.0 |
Selected to generate TypeDoc output as Markdown under docs/api. |
[WORK RULES]
- Do not write a plan or preamble; start working directly.
- Read in parallel; inspect independent files at the same time.
- Run tests after every meaningful set of changes.
- Do not claim the work is finished until tests pass.
- Update the dependency table in this file before adding a new library.
[ERROR PROTOCOL — 5 ATTEMPTS]
If an error occurs, follow this order:
- Read the full error, check the official documentation, try a different fix, and test.
- Use a completely different strategy from the first attempt, then test.
- Research an alternative method, then test.
- Isolate the issue, build a minimal reproducer, solve it layer by layer, then test.
- If necessary, try the most radically different approach, then test.
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 · 88 lines · 1,292 tokens per session scan A 9df7b98523fe
debug-recorder-mcp AGENTS.md is an instructions file published in the GitHub repository oaslananka/debug-recorder-mcp (2 stars, last pushed 28d ago), licensed MIT. It adds 1,292 tokens to every session, about $0.0065 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
health-monitor-mcp copilot-instructions.md
Copilot instructions for oaslananka/health-monitor-mcp, covering github copilot instructions, expectations, validation and output quality.
health-monitor-mcp AGENTS.md
AGENTS.md instructions for oaslananka/health-monitor-mcp, covering agents.md — ai agent instructions, tool-specific mirrors, environment check, registry reference and approved dependency pins.
health-monitor-mcp CLAUDE.md
Claude Code instructions for oaslananka/health-monitor-mcp, covering claude.md — repository instructions, role, startup order, repository standards and working rules.
silentfail CLAUDE.md
Claude Code instructions for decksaga/silentfail, covering silentfail, commands, stack, modes and key files.
health-monitor-mcp GEMINI.md
Gemini CLI instructions for oaslananka/health-monitor-mcp, covering gemini.md — repository instructions, role, startup order, repository standards and working rules.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).