Skip to content

falsiflow_public_issue_queue.md

Latest commit

 

History

History
98 lines (67 loc) · 3.67 KB

falsiflow_public_issue_queue.md

File metadata and controls

98 lines (67 loc) · 3.67 KB
Raw

Falsiflow Public Issue Queue

This page keeps the first public issue queue small, concrete, and easy to verify. Use it after launch posts start receiving feedback, or before launch to seed contributor-friendly work.

Queue Rules

  • Every issue should name one affected artifact: README, public demo, downstream PR, GitHub Action, starter template, adapter profile, or launch metrics.
  • Every issue should include a verification command or public URL.
  • good first issue means a newcomer can finish without changing the core claim engine.
  • help wanted means external experience would improve the outcome.
  • launch issues should be closed or updated during the 24-hour, 72-hour, 7-day, and 14-day launch review windows.
  • Do not request private evidence uploads in public issues.

Starter Labels

  • launch: public launch, demo, positioning, or adoption assets.
  • good first issue: small documentation, demo, fixture, or example work.
  • help wanted: useful external feedback or implementation help.
  • template: starter template or template release work.
  • evidence-gate: claim gate, evidence contract, or provenance behavior.
  • documentation: README, guide, troubleshooting, or launch copy.

Seed Issues

Create these as public issues before broad launch distribution:

  1. Capture a short downstream PR proof clip

    Labels: launch, documentation, good first issue

    Status: completed by docs/assets/falsiflow_downstream_pr_proof_strip.svg and tracked in issue #18.

    Goal: add a short GIF/video or screenshot sequence that shows the downstream PR moving from claim_check_blocked to claim_check_ready.

    Evidence: link the downstream PR, blocked run, ready run, and generated asset path.

  2. Add a local LLM eval quickstart note

    Labels: documentation, help wanted, evidence-gate

    Status: completed by docs/falsiflow_local_llm_eval.md and tracked in issue #19.

    Goal: show how Ollama, LM Studio, llama.cpp, or another local model runner can produce source-backed eval evidence without Falsiflow opening an API server.

    Evidence: falsiflow quickstart --template ai_claim_evaluation and falsiflow doctor --strict still pass.

  3. Promote the RAG quality gate proposal toward a starter template

    Labels: template, evidence-gate, help wanted

    Status: completed by the bundled rag_quality_gate starter template and tracked in issue #20.

    Goal: turn the documented RAG gate proposal into a bundled template only after the evidence contract, raw export fixture, placeholder blocker, and responsible-use boundary are clear.

    Evidence: template-check, casebook-check, and release-check pass.

  4. Add a product-metric downstream smoke example

    Labels: template, launch, good first issue

    Status: completed by examples/downstream_product_metric_smoke and tracked in issue #21.

    Goal: mirror the AI eval downstream story for a product metric claim such as activation lift, retention change, or conversion improvement.

    Evidence: the example has one blocked placeholder run and one ready source-backed run.

  5. Run the first launch metrics review

    Labels: launch, help wanted

    Goal: record the 24-hour launch snapshot after the first public post: stars, forks, install friction, repeated questions, demo clicks, and next follow-up issues.

    Evidence: update launch_metrics.md or a launch review issue with public post URLs and the review window.