agent-plugins

Find help for a bug, a small change, a difficult question or a handoff. Try one source-loaded task, see the result and its limits, then go deeper when useful. Choose from 24 task-method plugins; no architecture lesson is required to begin. Try the first task or inspect the core.

Choose an installation route or ask your agent to guide it

Try source or session loading first. Persistent installation needs an explicit scope and a complete package, then discovery and a read-only invocation.

Compare installation routes and limits · Copy a guided-install prompt.

Choose your next step

Start here

Find existing code

A source-loaded first task, actual findings, installation choices and undo.

Guides

Bug, feature, handoff

Follow real repository work, with evidence and boundaries at each stage.

Inspect the evidence

Evidence

Before & after examples

Paired examples with evidence labels and provenance. Some entries are ungraded seeds; inspect the source rather than assuming measured benefit.

Reference

How every change is checked

The full inventory of eval tiers, what each proves and what it structurally cannot, when it fires and what it costs — the doc the cheap tier itself keeps in sync with the workflows.

Deep dives 1 plugins with a long-form page

Deep dive

agent-compiler

Compile deterministic, content-hashed agents from small behavior modules: fuzzy intent becomes a typed AgentQuery, a stdlib-only kernel resolves modules and fails closed on conflicts, and out comes an immutable AgentImage with per-line provenance.

All plugins

  1. agent-compiler v0.3.1

    examplemodel eval pack

    Compose versioned behaviors into reproducible definitions with provenance and effect limits.

  2. codebase-design v1.0.0

    subtleno model pack

    Compare materially different interface designs before choosing one.

  3. consolidation-map v0.0.1

    no example yetno model pack

    Account for every entity in a consolidation and reconcile its disposition.

  4. context-handoff v1.0.0

    moderateno model pack

    Choose how to continue at a phase boundary and pass source references onward.

  5. delegated-intelligence v0.2.0

    no example yetno model pack

    Brief real workers, own outputs, integrate evidence and recover only invalid work.

  6. dev-diary v1.0.0

    moderateno model pack

    Discuss what mattered in the day's work and record or revise a journal entry.

  7. diagnosing-bugs v1.0.0

    exampleno model pack

    Test possible causes, make a focused fix and demonstrate regression coverage.

  8. docs-hygiene v1.0.0

    subtleno model pack

    Audit instruction files against current repository behavior and resolve contradictions.

  9. egress-gate v0.0.1

    exampleno model pack

    Inspect outbound content and its destination before transmission.

  10. eval-ladder v0.0.1

    no example yetno model pack

    Choose proportionate checks and describe their coverage gaps.

  11. find-before-build v0.0.1

    examplemodel eval pack

    Assess existing implementations before introducing another one.

  12. graveyard v1.0.0

    no example yetmodel eval pack

    Archive repositories as restorable bundles and prepare guarded deletion scripts.

  13. grill-me v1.0.0

    starkno model pack

    Examine a plan through questions scaled to uncertainty and stakes.

  14. plugin-factory v1.1.0

    moderateno model pack

    Scaffold a plugin with manifests, entry points and checks that start red.

  15. recurrence-detector v0.0.1

    moderateno model pack

    Propose improvements from repeated sourced failure patterns.

  16. scope-fence v0.0.1

    examplemodel eval pack

    Keep diffs tied to the agreed task and record unrelated findings.

  17. semver-gate v1.1.0

    examplemodel eval pack

    Classify action risk and rehearse restoration before irreversible changes.

  18. stop-rule v0.0.1

    examplemodel eval pack

    Bound repeated attempts and return an evidence-backed stop report.

  19. tailscale-wif v1.0.0

    no example yetmodel eval pack

    Set up secretless GitHub Actions authentication to Tailscale.

  20. tracer-bullets v1.0.0

    moderateno model pack

    Build one real end-to-end slice and expand it.

  21. verify-before-claim v1.1.0

    no example yetmodel eval pack

    Check completion evidence; optionally establish and verify executable acceptance criteria.

  22. voice v0.1.0

    no example yetmodel eval pack

    Improve prose and machine-facing output; includes explicitly requested second-opinion review.

  23. wave-research v0.1.0

    no example yetno model pack

    Examine a contested claim through rival explanations and sourced adversarial review.

  24. wayfinder v1.0.0

    examplemodel eval pack

    Map a larger effort into decisions and dependent work across sessions.

This site is generated, and checked. The landing page and the example gallery are rendered from committed data by a script in docs/; the offline eval tier fails any commit whose page drifts from its data, and the Pages workflow re-checks before it deploys.