💻 WorkNew

46. Audit a site's agent-readiness

The job: Give Muse a site URL — it checks whether an autonomous agent can actually use the site: discovery files, machine-readable endpoints, onboarding friction, and ID consistency, ranked by where an agent would give up.

Audit [site URL] for how well an autonomous AI agent can use it. Steps: 1) Discover: fetch /agents.md, /agent.json, /llms.txt, /robots.txt — note which exist and whether an unaided agent could find the docs. 2) Inventory every machine-readable endpoint (RSS/Atom, sitemap.xml, public JSON APIs) and test each with a real request: record HTTP status, latency, and auth requirements. 3) If there's an agent onboarding flow (keypair/API key/OAuth), walk it end-to-end and note every friction point. 4) Check schema consistency: do IDs/slugs match across endpoints? Deliver a table: endpoint | method | auth | status | latency | notes, plus the top 3 friction points ranked by how likely they are to make an agent give up. Do not submit, post, or mutate anything during the audit — read-only until I approve.

Difficulty Moderate · Best for Site owners who want agent traffic but have never watched an agent try to use their site · Developer advocates launching a docs page, API, or agent onboarding flow · Anyone adding /agents.md or /llms.txt and wondering if it's enough

What Muse will do:

Give Muse the site URL and keep the audit read-only
It fetches the discovery files (/agents.md, /agent.json, /llms.txt, /robots.txt)
It inventories and test-calls every machine-readable endpoint, recording status, latency, and auth
It walks any agent onboarding flow end-to-end and notes the friction
You get the endpoint table plus the top 3 abandonment risks to fix first

Why it works: Humans forgive maze-like docs and inconsistent slugs; agents often just stop. This audit reads your site the way an autonomous user would: can it find the docs, call the endpoints, and get through signup without a human rescue? The deliverable is ranked by abandonment risk, so you fix the places agents bail first instead of polishing pages they never reach.

Common failures:

A linked endpoint isn't a working endpoint — test auth, status, and latency, don't trust the docs
Keep it read-only: no signups that mutate state, no test posts, until you approve
Latency is a snapshot, not a verdict — re-run critical endpoints before declaring them broken

Prompt variations:

Run the same rubric on a competitor's site and compare scores
Re-run after shipping docs or API changes to track readiness over time
Scope it down: audit only the public docs surface, no account signup

Submitted by PM-Evaluator

Was this useful?

Discussion

Loading comments…