How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

Check Social Content Accuracy - Prompt

Name: Check Social Content Accuracy
Author: guysolan

Prompt file imported from guysolan/thetis (.cursor/commands/check-social.md). Copyright stays with the author.

Check Social Content for Medical Accuracy

Review the draft the user provides (pasted text or a file path) for medical accuracy.

Steps

1. Load the governance rules

Read these files — they are the single source of truth and always win over any other content:

.cursor/rules/achilles-clinical-positions.mdc (30 canonical clinical positions)
.cursor/rules/content-guidelines.mdc (voice + safety rules)

2. Find related published content

For each clinical claim in the draft, search the canonical content sources:

apps/course/src/content/course/ — course lessons (deepest clinical detail)
apps/website/src/content/guide/ — week/stage guides
apps/website/src/content/blog/ + apps/website/src/content/articles/ — published articles
apps/website/src/sections/FAQs/ — FAQ answers
apps/achilles-rupture/src/content/articles/ — educational articles

Use semantic search and grep for key terms (e.g. "wedge removal", "blood thinners", "return to sport").

3. Review

Check the draft for:

Claims contradicting any clinical position (severity: high)
Claims contradicting published content (severity: medium — flag which source)
Missing warning signs / safety info where the topic touches risk
Tone violations (see content guidelines: plain language, no false precision, honest about uncertainty)
Invented statistics or facts not present in any source

4. Output

## Verdict: PASS | REVISE | BLOCK

## Summary
One sentence.

## Issues
- [HIGH|MEDIUM|LOW] "exact claim text"
  Problem: what is wrong, citing the position number or source file path
  Fix: suggested replacement

## Missing safety info
- ...

## Supporting sources
- file paths that back up the draft's correct claims

BLOCK if any claim directly contradicts a clinical position or gives dangerous advice. Note: the same check is available with vector retrieval in the chat app (apps/chat, /check route).

Install Check Social Content Accuracy and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "prompt": "# Check Social Content for Medical Accuracy\n\nReview the draft the user provides (pasted text or a file path) for medical accuracy.\n\n## Steps\n\n### 1. Load the governance rules\n\nRead these files — they are the single source of truth and always win over any other content:\n\n- `.cursor/rules/achilles-clinical-positions.mdc` (30 canonical clinical positions)\n- `.cursor/rules/content-guidelines.mdc` (voice + safety rules)\n\n### 2. Find related published content\n\nFor each clinical claim in the draft, search the canonical content sources:\n\n- `apps/course/src/content/course/` — course lessons (deepest clinical detail)\n- `apps/website/src/content/guide/` — week/stage guides\n- `apps/website/src/content/blog/` + `apps/website/src/content/articles/` — published articles\n- `apps/website/src/sections/FAQs/` — FAQ answers\n- `apps/achilles-rupture/src/content/articles/` — educational articles\n\nUse semantic search and grep for key terms (e.g. \"wedge removal\", \"blood thinners\", \"return to sport\").\n\n### 3. Review\n\nCheck the draft for:\n\n1. Claims contradicting any clinical position (severity: high)\n2. Claims contradicting published content (severity: medium — flag which source)\n3. Missing warning signs / safety info where the topic touches risk\n4. Tone violations (see content guidelines: plain language, no false precision, honest about uncertainty)\n5. Invented statistics or facts not present in any source\n\n### 4. Output\n\n```\n## Verdict: PASS | REVISE | BLOCK\n\n## Summary\nOne sentence.\n\n## Issues\n- [HIGH|MEDIUM|LOW] \"exact claim text\"\n Problem: what is wrong, citing the position number or source file path\n Fix: suggested replacement\n\n## Missing safety info\n- ...\n\n## Supporting sources\n- file paths that back up the draft's correct claims\n```\n\nBLOCK if any claim directly contradicts a clinical position or gives dangerous advice. Note: the same check is available with vector retrieval in the chat app (`apps/chat`, /check route).", "variables": [], "notes": "Review a draft social post or article for medical accuracy against the clinical positions and all published content in the monorepo. Tools: Read, Grep, Codebase.", "metadata": { "source": { "provider": "github-cursor-commands", "repository": "https://github.com/guysolan/thetis", "path": ".cursor/commands/check-social.md", "ref": "e51327d3878325917136ecfd5c01f7d4e9e560b6", "url": "https://github.com/guysolan/thetis/blob/e51327d3878325917136ecfd5c01f7d4e9e560b6/.cursor/commands/check-social.md", "key": "guysolan/thetis/.cursor/commands/check-social.md" } } }

Check Social Content Accuracy

About

Check Social Content for Medical Accuracy

Steps

1. Load the governance rules

2. Find related published content

3. Review

4. Output

Install Check Social Content Accuracy and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Use it

Manifest

Reviews