Collect
Crawls docs plus README, AGENTS.md, LLMs.txt, quickstarts, and optional GitHub/OpenAPI sources.
Monitor Your Docs
Mantis crawls your docs, README, LLMs.txt, quickstarts, API specs, and repo files, then validates snippets and cross-checks claims before docs drift reaches production.
Sources Monitored
0
Snippets Verified
0
Cross-Doc Conflicts
0
Last Consistency Score
0
What It Does
Crawls docs plus README, AGENTS.md, LLMs.txt, quickstarts, and optional GitHub/OpenAPI sources.
Runs syntax/import/runtime checks and replacement suggestions for missing or deprecated packages.
Applies DocPrism categories with optional LLM + Smart/Full RnR and constrained evidence filtering.
Ships CI/CD and CLI parity with conflict views (snippets, evidence, timeline, charts) to block stale releases.
How It Works
Add docs URL + optional repo and OpenAPI schema for ground-truth consistency checks.
Mantis crawls, validates, extracts entities, and runs DocPrism/RnR analysis.
Inspect inline diff, linked evidence cards, conflicts tab, and staleness timeline.
Use GitHub Action + CLI exit codes to fail builds before stale docs reach users.
Interactive Demo
Package and version staleness
Detects unresolvable packages and deprecated versions.
Broken CLI and API commands
Validates shell examples, flags incorrect flags and endpoint paths.
Cross-document contradictions
Flags README/API/quickstart disagreements with evidence pairs and categorization.
Timeline and analytics drift
Tracks scan volume, error trends, language distribution, and entity drift across runs.
Integrations
Analytics + Evidence
Scan Volume
Error Trend
-23%
Compared to previous 30 days
Top Issues
Snippets View
Severity + category filters with on-demand LLM evidence.
Conflicts View
Cross-doc contradictions with linked sources and explanations.
Timeline View
Entity-level drift tracking across scans and releases.
Why Mantis
| Capability | Mantis | DIY Claude / Script |
|---|---|---|
| Deterministic checks | Built-in (syntax, imports, registry) | Manual / not included |
| Token cost per run | Batched + cached, small models first | Full cost every invocation |
| CI/CD integration | One command, exit codes, PR comments | Custom scripts to maintain |
| Setup time | Under 2 minutes | Hours to days |
| Cross-doc consistency | Automatic entity extraction + conflict detection | N/A or manual comparison |
| Caching | Hash-based, skip unchanged snippets | Build your own |
| Repeatable results | Same input, same pass/fail, every run | Prompt variance across runs |
A senior dev can build this. We're selling "don't build it, run it."
Beta Program
Help us build the doc quality gate your team actually uses. Ship with confidence that your docs match your code.
No credit card · CLI ships today · Cancel anytime
Building with