Replay-mode managed-agents for engineering teams

Replay-mode managed-agents for engineering teams. CI, code review, incident triage. Part of the subagentdispatch fleet — same chassis, same replay layer, vertical-specific corpus. (29+ vendor signals indexed across the fleet.)

Browse data Try the demo API endpoint →

Replaying · 0 API calls · 0 tokens

Ticket-deflector — Triage a refund-after-cancel support ticket

claude-haiku-4-5

user ▸ Classify this support ticket and draft a reply if it's auto-respondable: 'I cancelled my Starter plan yesterday but got charged again for $29.'

Classification: (b) escalate to billing — duplicate-charge after a cancellation requires a finance review.

Draft reply (held for billing review, not sent):
Thanks for flagging this — I'm pulling the cancellation and the new charge now so billing can confirm both in one pass. You'll get a refund + a written confirmation that the Starter plan is fully closed within one business day.

Driven by lib/replay/web-data.ts · same chunks the CI replay layer streams in npm run demo:finance:earnings.

0+

Active sources

0h

Last fetched

AI-powered

Intelligence analysis

Native

MCP server

Platform capabilities

More than a feed — a complete intelligence layer for vendor documentation.

Intelligence analysis

AI-powered analysis of vendor documentation changes. Get significance scoring, categorization, and key quotes extracted from every meaningful update.

MCP server

Connect directly from Claude Desktop, Cursor, or any MCP-compatible AI tool. Your AI assistant can query vendor docs natively.

29+ vendor sources

Comprehensive coverage across AI platforms, developer tools, databases, payments, and cloud providers. New sources added continuously.

Trusted by developers

Real-time intelligence powering AI-native workflows.

0+

Vendors tracked

Across AI platforms, dev tools, databases, payments, and cloud providers

0x daily

Diffs detected

Every 6 hours, content changes are identified, versioned, and analyzed

REST + MCP

API requests served

JSON API, webhook delivery, and native MCP server for AI tools

Tracked vendors

29+ active sources — fetched and versioned every 6 hours.

AnthropicAI Platform

https://platform.claude.com/llms.txt

+ llms-full.txt available

Vercel AI SDKDeveloper Tools

https://ai-sdk.dev/llms.txt

https://neon.com/llms.txt

https://docs.stripe.com/llms.txt

SupabaseDatabase

https://supabase.com/llms.txt

LangChainDeveloper Tools

https://docs.langchain.com/llms.txt

PineconeDatabase

https://docs.pinecone.io/llms.txt

OpenAIAI Platform

https://developers.openai.com/api/docs/llms.txt

+ llms-full.txt available

Google GeminiAI Platform

https://ai.google.dev/gemini-api/docs/llms.txt

+ llms-full.txt available

LinearDeveloper Tools

https://linear.app/llms.txt

NotionDeveloper Tools

https://developers.notion.com/llms.txt

+ llms-full.txt available

https://api.slack.com/llms.txt

CloudflareCloud

https://developers.cloudflare.com/llms.txt

GitHubDeveloper Tools

https://docs.github.com/llms.txt

DatadogDeveloper Tools

https://docs.datadoghq.com/llms.txt

https://www.twilio.com/llms.txt

https://resend.com/llms.txt

CursorDeveloper Tools

https://cursor.com/llms.txt

https://vercel.com/llms.txt

DockerDeveloper Tools

https://docs.docker.com/llms.txt

MongoDBDatabase

https://www.mongodb.com/llms.txt

https://redis.io/llms.txt

https://docs.snowflake.com/llms.txt

https://docs.databricks.com/llms.txt

https://docs.confluent.io/llms.txt

+ llms-full.txt available

https://turso.tech/llms.txt

BrowserbaseDeveloper Tools

https://www.browserbase.com/llms.txt

FirecrawlDeveloper Tools

https://www.firecrawl.dev/llms.txt

+ llms-full.txt available

PagerDutyDeveloper Tools

https://www.pagerduty.com/llms.txt

How it works

Three steps from raw vendor docs to actionable intelligence.

[~]

1

We crawl llms.txt files

Every 6 hours, we pull llms.txt from 29+ vendors, compute SHA-256 content hashes, and store versioned snapshots.

Sanitized + scanned for injection

</>

2

We detect and analyze changes

When content changes, AI-powered intelligence kicks in — significance scoring, categorization, and key quote extraction.

Line-level diffs + link tracking

{=}

3

You get intelligence via API

Structured JSON API, webhook notifications, and native MCP server access. Query from your AI tools or integrate via REST.

REST + MCP + Webhooks

See what's changing right now

Search any tracked vendor to see their latest documentation changes, diff summaries, and intelligence.

Try: