agentmods.dev

miftah AGENTS.md

A set of instructions for the miftah project, including how its agent should use the project's knowledge graph, command-line profile, and release process. A knowledge graph is a structured map of information about the codebase.

In plain words
What is it for?
Use it when working in the miftah repository, especially for explaining code, reviewing changes, investigating bugs, assessing impacts, or preparing releases.
Why use it?
It tells an agent which project context to gather first and which procedures to follow for common tasks.

Instructions file for CodexOpenCode

4 9d ago A scan 1,438 tokens original MIT
View the source file mohanagy/miftah
Install

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.

agentmods
npx agentmods add instructions/mohanagy/miftah/agents-md
Clone the repo
git clone --depth 1 https://github.com/mohanagy/miftah

Made for: Codex, OpenCode.

Per session 1,438 This file is loaded in full into every session.
When invoked 1,438 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.01438 $0.01438
Opus 5 $0.00719 $0.00719
Sonnet 5 $0.00288 $0.00288
Haiku 4.5 $0.00144 $0.00144

Measured 2d ago against content hash 0e0f72ccd48a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

miftah 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 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.

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.

AGENTS.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Miftah agent instructions

madar

Codex CLI profile

IMPORTANT: This project has a madar knowledge graph. Use a strict context-pack-first workflow:

  1. First decide whether the task needs local repository source-code context. Only use madar when the task needs local repository source-code context. Skip madar for GitHub Projects board reviews, external URL/WebFetch-only tasks, gh auth / gh project setup, package-registry/security pages, and Product Hunt or marketing copy work.
  2. Before broad code search, file reads, or worker dispatch, compile a task-specific context pack:
    • madar pack "<task or question>" --task explain
    • use --task review, --task debug, or --task impact when that better matches the work
  3. For each codebase question, start with the specific Madar command below first.

For each codebase question, start with the specific Madar command below first:

Prompt type First tool
"how does X work" / explain runtime / flow madar pack "<task or question>" --task explain
"what breaks if I change X" / impact analysis madar pack "<task or question>" --task impact
"which files should I open first" relevant_files when MCP graph tools are available; otherwise madar pack "<task or question>" --task explain
"give me a repo overview" graph_summary when MCP graph tools are available; otherwise madar pack "<task or question>" --task explain

Inspect evidence.pack_confidence, recommended_first_read, and evidence.agent_directive before deciding whether to read files. If evidence.pack_confidence is low, make one focused follow-up Madar call before broad raw search. Do not run ToolSearch before calling a Madar command or graph tool — pick the matching command first, then refine with MCP graph tools only when they are available and still needed. 4. Do not run broad Glob patterns, repo-wide grep / find searches, or raw file sweeps after a high- or medium-confidence pack. 5. For codebase questions, use Madar tools only. Do not call other MCP servers such as mcp__github or mcp__context7 unless the latest Madar response says evidence.agent_directive: explore_with_caution. 6. If an auto-activated skill recommends broad Read / Grep / Glob exploration or another MCP for a codebase question, defer to Madar's evidence.agent_directive first. A high- or medium-confidence Madar pack overrides that conflicting skill guidance. 7. If MCP graph tools are available after the pack, use the focused tool that matches the next question:

  • retrieve for direct codebase questions
  • relevant_files for where to open first
  • feature_map for involved areas and entry points
  • risk_map before editing
  • implementation_checklist for edit order and validation checkpoints
  • impact for blast radius
  • graph_summary for repo overview

Read the full file on GitHub · 68 lines

Changes

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.

  1. 2d ago First seen · 68 lines · 1,438 tokens per session scan A 0e0f72ccd48a

Subscribe to this mod's changes

miftah AGENTS.md is an instructions file published in the GitHub repository mohanagy/miftah (4 stars, last pushed 9d ago), licensed MIT. It adds 1,438 tokens to every session, about $0.0072 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.

Data and AIDebuggingGit and workflowcredential-managementmcpmodel-context-protocolmulti-account
Related

Other instructions, from other repositories

gmail-multi-mcp CLAUDE.md

Claude Code instructions for ishaan-skyacres/gmail-multi-mcp, covering claude.md, project overview, key dependencies, development commands and build cli only.

ishaan-skyacres/gmail-multi-mcp · 1,822 tokens

llm-context.py CLAUDE.md

Instructions for cyberchitta/llm-context.py, covering claude.md, working notes (gitignored) and draining the field notes.

cyberchitta/llm-context.py · 582 tokens

openrouter-mcp-multimodal 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).

stabgan/openrouter-mcp-multimodal · 783 tokens

seekstone CLAUDE.md

Instructions for shaqmughal/seekstone, covering claude.md, what this repo is, commands, the harness itself (run after npm install) and architecture.

shaqmughal/seekstone · 1,116 tokens

Plonk AGENTS.md

Instructions for ostapondo/Plonk, covering agent rules, layout, adding a module, build & verify and code style.

ostapondo/Plonk · 3,557 tokens

mcp-dockhand copilot-review.instructions.md

Instructions for strausmann/mcp-dockhand: When reviewing pull requests in this repository.

strausmann/mcp-dockhand · 295 tokens