Guides
Short, practical how-tos for real tasks. Each one is copy-pasteable end to end. New to the tool? Start with the getting-started tour; come here when you have a specific job to do.
Set a goal contract
State the goal as a contract — accomplish these things (typed acceptance criteria over your code), subject to these policies (the rigor bar). Then read back met / governed / certified.
AuthorCapture & curate a trace
You just finished an agent session — turn it into a clean, honest reasoning record: emit, curate the narrative, declare the gaps.
ReviewerReview an AI-generated PR
Someone handed you an AI-written PR. Confirm the goal was met and governed, then certify the criteria were the right ones — the certified axis is the reviewer's job.
MaintainerAdd ponens to CI / your PRs
Wire up the GitHub Action so every PR reads back met / governed / certified, with the governed axis as a blocking PASS/FAIL gate and a one-click viewer.
MaintainerGovern a repo with policies
Own the governed axis: pull best-practice policies from the gallery, scope them to goals, block by default (disable or waive on the record), and write your own rules in LTLf.
IntegratorUse ponens programmatically
Drive the policy checker from your own tool: machine-readable JSON output, the Python evaluator API, and a live per-change compliance loop.
IntegratorImport from your observability stack
Turn your existing OpenTelemetry or Langfuse agent traces into ponens traces — spans/observations → actions + lineage — then check them against policies.