tessl-labs/learn-from-past-work
Skills that Learn from past work loops run: review pull request history for agent context and Tessl verifier improvements, read recent agent runs to record recurring problems as project findings, propose Code Review lens changes from how the lenses did on recent pull requests, and turn open findings into tickets the factory builds.
94%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Medium
Suggest reviewing before use
README.md
Learn from past work
The skills that let a Tessl factory learn from its own history. On a schedule, Tessl reads recent pull requests, their reviews, and agent runs, records the problems that keep coming back as findings, and can turn those findings into tickets the factory builds.
You don't install this plugin yourself. Learn from past work runs it once it's turned on for a project, in the Tessl web app or with the tessl-labs/factory-setup plugin.
Skills
- learn-from-review-feedback: find mistakes reviewers keep correcting, and propose agent context or checks that prevent them.
- learn-from-agent-runs: find problems that recur across agent runs, such as dead ends, missing access, or flaky checks.
- tune-review-lenses: propose changes to Code Review from the findings people addressed, declined, or missed.
- findings-triage: check open findings against the current code, and file tickets for the ones worth building.