Orchestrate automated code scanning, structured review, and verified auto-fixing for repositories.
npx skills add https://github.com/tcytan/tecton --skill team-review-tcytan
Please help me install this Agent Skill. Skill: team-review Source: https://github.com/tcytan/tecton/tree/main/.claude/skills/team-review Command: npx skills add https://github.com/tcytan/tecton --skill team-review-tcytan
It reduces the time and inconsistency of code review by running a structured, multi-stage review pipeline that scans, deep-triages findings, and can apply fixes with verification and rollback.
Run team-review on your target codebase (e.g., the repository folder) and ask it to perform a quick scan and then continue with review if findings are found.
High-intent search queries and answers about installing and using this skill.
You can automate code review by running a structured pipeline that scans your repository, deep-triages findings across security, correctness, performance, and maintainability dimensions, and optionally applies verified fixes with rollback capabilities.
Rollback-on-failure works by applying changes in dependency-aware groups and running verification tools afterward. If verification fails, the automated fixing process reverts the failed changes to keep your workspace stable.
Yes, you can run a quick scan-only code analysis by using the quick mode. This executes the scanner to identify findings without proceeding to the deep-triage review or auto-fix stages, allowing you to inspect results first.
The best way to coordinate multi-agent static analysis is using phase-based task dependencies to pass standardized findings schemas between the scanner, reviewer, and fixer. This ensures deterministic plan outputs and tool-based verification before applying changes.
Yes, the automated code review pipeline supports a full mode with a user confirmation checkpoint. This pauses execution after the review phase, allowing you to approve findings before the fixer applies automated changes.
Yes, you can apply fixes from an existing manifest without re-scanning by using the fix-only mode. This skips the scanner and reviewer stages, directly executing the fixer with rollback-on-failure verification based on your provided manifest.
utsavthakurAudit codebases with static analysis and generate structured risk reports.
Khang5687Monitor ClawSweeper workflow status via GitHub API events.
raphaelchristiRun repeated evaluations and compute score stability statistics for evolved LLM agents.
thktOrchestrate AI agents to audit code quality with tiered review and static analysis.
