Catch broken paths and commands before your next merge.
70/70 references verified · edf5493
CLAUDE.md:7
npm run build
CLAUDE.md:7
npm run clean && tsc
CLAUDE.md:8
npm run watch
Claude Code, Cursor and Copilot can use the context files in your repo. A PR that changes a path or command can leave those instructions behind.
Proposed on a branch you review
The proposed change is reviewable before it is merged.
Each recognized reference gets one of three states
Recognized commands are matched to manifests and paths to the file tree. This deterministic check does not assess all prose or settle what the author meant.
The audit runs on its own and costs nothing. The correction is the part you ask for.
Install the GitHub App or add your GitLab project. The first audit runs in about a minute and tells you what your context files get wrong today: supported commands and paths they claim, resolved against the code at HEAD.
Nothing to add to your pipeline. Kontrua posts a kontrua/context commit status and one comment naming each broken claim with its file, line and reason. A claim it cannot decide is reported, never failed, so the check stays worth requiring.
The correction lane opens a pull request that repairs the drifted lines: the context artifacts Kontrua owns, plus single-line corrections to the files you wrote. You review it like any other PR. This is the part that is metered.
The audit needs no trigger: every push and every PR gets its verdict, free. These are the ways to launch a correction, the run that rewrites what broke.
Install the GitHub App or connect your GitLab project. Kontrua audits every pull request for free, and when the repository is set to trigger on approval it runs the correction for the claims that pull request broke.
Trigger mode is per repository: on PR approval, on merge, or manual only.
Nothing to paste
Kontrua receives the pull request event from the GitHub App or your GitLab project, so both the audit and the correction start without a snippet in your repository.
Hand-maintained context vs. Kontrua.
Audits and PR verdicts are unlimited on every plan: supported command and path references are checked on connection, push and pull requests. Files and prose outside coverage are disclosed. The allowance counts correction runs; the API trigger and Notion sync work on every plan, while scheduling needs a paid one. Open-source repos get the published open-source allowance.
For solo devs and small projects
or $24 per developer per month, billed annually
For growing engineering teams
or $48 per developer per month, billed annually
For teams shipping at scale
Custom solutions for your team
Kontrua checks supported command and path references in selected instruction and documentation files against the repository at a commit. It reports files with no supported checks and files outside scope. It does not judge every sentence or every nested file.
The verdict uses deterministic checks against manifests and the file tree, without a model. A supported reference can pass, fail, or be reported as unverifiable. A deterministic result can still miss context or be wrong about what the author meant, so review the evidence.
The audit reads existing files without editing them. Existing files are audit only by default. You can opt specific paths into targeted correction proposals or managed regeneration. Missing enabled outputs can be generated. Review the file diff before merging.
An existing AGENTS.md stays team-owned by default. Kontrua can audit supported references; it proposes edits only if you explicitly permit corrections for that path or delegate the file for managed regeneration. Review the full diff before merging.
Any tool that reads context files from your repo: Claude Code, Cursor, Copilot, and anything else following the AGENTS.md convention. Standard files, no SDK, no lock-in.
Audits and PR verdicts are free and unlimited on every plan, including Free. You pay per developer for corrections, the runs that rewrite what broke.
The audit reads repository contents and records findings and run metadata. Correction runs can send selected repository context to the model provider and store proposed outputs. Changes are reviewable in a branch or an existing pull request where that integration is configured. See the privacy policy for the data flow.
Connect a repo to check supported command and path references. The result shows what passed, what broke, and what could not be checked.