- Home
- Skills
- Prompt & Skill Engineering
- agent readiness audit
Works with the AI Agents you already use
agent readiness audit
Audit your repository for AI agent compatibility and generate prioritized fixes to improve coding assistant performance.
Free
agent readiness audit
Example session with this skill installed
Audit this repo for agent-readiness and generate an AGENTS.md starter if needed.
- Read your context and instructions
- Compiled the agent readiness audit
I have audited the repository and identified critical gaps in agent orientation and the verification loop. While the project structure is clean, the lack of explicit command mappings and large file sizes are causing agent friction.
Agent Readiness Report — frontend-service — 55/100
| Dimension | Score | Max | Key evidence |
|---|---|---|---|
| Instructions | 0 | 25 | No .cursorrules or CLAUDE.md found in root. |
| Verify loop | 10 | 20 | package.json has scripts, but npm test fails due to missing env. |
| Orientation | 15 | 15 | Excellent README.md with clear architecture diagrams. |
| Navigability | 10 | 15 | 4 files exceed 800 lines (e.g., ApiClient.ts). |
| Tests | 15 | 15 | 92% coverage on critical paths via Jest. |
| Guardrails | 5 | 10 | GitHub Actions present, but no branch protection. |
Top issues (impact × effort)
Missing Instruction Files — Agents are guessing style and commands. Create CLAUDE.md with the discovered scripts below. Effort: S.
2.
Brittle Test Environment — npm test requires a local Redis instance not mentioned in docs, causing agent loops. Effort: M.
Quick wins (≤30 min each)
- Add a
.cursorrulesfile to ignore thedist/andbuild/directories to save context. - Split
ApiClient.tsinto smaller service modules to stay under 500-line limits.
AGENTS.md starter
# Agent Instructions
## Build & Test
- Build: `npm run build`
- Test: `npm test`
- Lint: `npm run lint`
## Project Structure
- `/src/components`: UI components (React/Tailwind)
- `/src/services`: API logic
- `/tests`: Integration and unit tests
## Standards
- Use TypeScript strict mode.
- Follow functional patterns; avoid classes where possible.
Next steps
- Commit the
AGENTS.mdstarter to the root directory. - Refactor
ApiClient.tsto improve navigability scores.
Connects securely to your tools. The creator never sees your data.
About this skill
When AI coding assistants or agents struggle to navigate your codebase, fail to run tests, or hallucinate architecture patterns, the project lacks agent-readiness. This skill audits your repository to diagnose exactly why agents are underperforming. It evaluates the project across six technical dimensions to produce a prioritized improvement plan.
What it does
- Instruction Audit checks for the presence and accuracy of
.cursorrules,CLAUDE.md, orAGENTS.mdfiles. - Verify Loop Analysis evaluates if the install-build-test cycle is discoverable and fast enough for agent iteration.
- Navigability Assessment identifies context-heavy files over 500 lines and identifies dead code that confuses LLMs.
- Guardrail Review checks for CI integration and protections that prevent common agent-generated errors.
- Scoring Engine assigns a 0 to 100 readiness score based on evidence-backed criteria.
How it works
- The user triggers the audit on a specific repository or directory.
- The agent performs a read-only scan of manifests, documentation, and directory structures.
- The agent runs safe, dry-run commands to verify scripts and build paths.
- It generates a markdown report with a score breakdown, top issues, and quick wins.
Frameworks & tools
This tool works with any repository containing standard manifests like package.json, requirements.txt, or pyproject.toml. It specifically looks for configuration files for Cursor, Claude Dev, and GitHub Actions.
Why this beats prompting it yourself
Writing a custom prompt for a repo audit often leads to vague advice like "add more comments." This skill uses a weighted 100-point scoring system that prioritizes the verify loop and machine-readable instructions, ensuring you fix the technical blockers that actually stop agents from shipping code.
Use cases
- Onboarding agents to a legacy codebase that lacks structured documentation.
- Optimizing Cursor/Claude performance by identifying context sinks and stale files.
- Setting up new repos with an
AGENTS.mdstarter based on discovered commands.
Known limitations
Does not automatically apply code fixes or refactors. Large repo scans are limited to safe, fast, dry-run commands and directory listing. Requires manual confirmation before running full install or test suites.
Read-only by default. Requires manual confirmation for expensive build/test commands. Scores unknown dimensions as zero rather than guessing.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 23 days ago
- Free with an Agensi account