botscoutKnow what you are reading.
Each result identifies the source, review date when available, and brief depth. Source review is not hands-on testing. Buyer ideas and offers are hypotheses.
FOR THE HUMAN. AND THEIR AGENT.
Your next project,
with the context attached.
Let your agent do the first pass. Every published find has structured data, original sources, and a portable research brief.
Free public readsNo API keySame catalog as the site
ONE PROMPT TO START
Give your agent a useful brief.
Copy this into an agent that can browse or make HTTP requests. It can help you compare options before you build.
Use https://botscout.app/llms.txt to explore BotScout. Ask about my task, technical comfort, and budget. Find up to three relevant projects using the public API. Read their Markdown briefs and original sources. Compare proposed outputs, setup, costs, evidence, and unknowns; cite both source and BotScout links. Recommend a small pilot and acceptance tests. Do not treat source claims as verified results or take actions on my behalf.
Open the agent guide ↗A small API. Useful answers.
OpenAPI schema ↗TRY A REAL REQUEST
GET /api/v1/discoveries?collection=no-code&limit=3Returns matching briefs with source evidence, practical facets, and links to JSON, Markdown, and the human-readable page.
| Read | Endpoint | You get |
|---|---|---|
| Find projects | /api/v1/discoveries | Filtered results and a next-page URL |
| Get a brief | /api/v1/discoveries/{id} | Structured context and source provenance |
| Take it with you | /api/v1/discoveries/{id}/brief.md | A complete Markdown research brief |
Available filters & pagination
q- Search title, summary, tool, and suggested buyer or deliverable. All words must match.
category- Exact category; omit for all. All discoveries · Work & productivity · Sales & marketing · Customer support · Research & data · Coding & development
tool- Exact tool name; omit for all.
collection- A guided collection; these are filters, not performance rankings. no-code · client-service · self-hosted
kind- Editorial playbook, scouted project, or community find. curated · discovered · community
sort- Recommended means featured/editorial first, not a performance score. recommended · newest · az
period- Added in the past seven days, not a source verification date. week
mode- Build style no-code · low-code · developer · unknown
hosting- Hosting option hosted · self-hosted · either · unknown
effort- Estimated first-pilot scope small · connected · custom · unknown
depth- Detail available in the brief full · starter
limit- Number of results; at most 50.
offset- Zero-based position in the filtered result set.
EVIDENCE THAT TRAVELS WITH ITS LIMITATIONS
Inspect the result behind the claim.
Read /api/v1/discoveries/{id}/evidence for published reports, then /api/v1/evidence/{id} for conditions, metrics, and version history. Preserve self-reported and execution-unverified labels.
Research before action.
This API only reads published finds. It cannot send messages, charge a card, or change a project. Ask for human approval before acting on an idea.
Keep requests considerate.
Up to 50 results per page and 120 requests per minute per network address. Honor Retry-After on 429 responses. GET and HEAD support ETag revalidation. Cite BotScout and the original source.
Access requires a tool that can read public URLs; this is not an installed integration or an MCP server. Markdown and JSON responses allow cross-origin public reads. Avoid personal information in search queries. llms.txt is a discovery aid, not a promise that an AI service will index or recommend BotScout.