Run cold-user rehearsals across legitimate inputs and outcomes
Nobody has claimed this yet.
- Dominant language
- No language data
- Stars
- 0
- Forks
- 0
- Avg merge
- 8h 21m
- Merged PRs (30d)
- 2
Description
Outcome
Cold users can supply their own supported trip and mandate and receive correct factual graph/policy outcomes without facilitator steering or hidden golden values.
Scope
- Vary routes, dates, hard constraints, soft preferences, USD caps, thresholds, and allowed actions.
- Include autonomous, approval, prohibited, no-match, temporary-unavailability, stale-selection/refusal, partial-success, and non-USD rejection cases.
- Verify the agent ranks/explains while Farewright only evaluates typed rules and enforces at execution.
- Exercise the full disruption/service/settlement loop only when the user's selected action passes execution guards.
- Ask hostile stakeholder questions about simulation, fallbacks, monitoring, PII, authority, and refunds.
Acceptance
- Facilitator explains only supported/test boundaries and does not steer request, budget, threshold, complaint, or outcome.
- The private reliability matrix is used for QA measurement only, not user selection.
- At least one honest no-match and one policy-prohibited/refused run end without being reframed as failure.
- A complete search may stop honestly; any optional recovery/service/settlement drill that proceeds uses the actual selected facts.
- A model attempt to reuse stale
autonomousevidence, supply its own budget math, or book one leg of an over-cap plan is rejected. - Measurements and failures are recorded without turning them into a deterministic script.
Dependencies
- #31 live-test/outage rehearsal and #34 private reliability artifact.
- All earlier milestone gates satisfied or explicitly documented.
Review
Inputs to the mandatory independent demo-integrity review.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reviewing the prerequisites in #31 and #34 and confirming that all earlier milestone gates are satisfied or documented. Use the issue's acceptance checklist to run varied cold-user rehearsals, record measurements and failures without a deterministic script, and provide the resulting material for the independent demo-integrity review.
Written by the indexing model from the issue text.
Assessment
- Domain
- testing-qa
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100