[adoption][validation] Validate newcomer comprehension, accessibility and failure recovery

[RobLe3]

Purpose

Test whether first-time users can understand and use the public IICP material without relying on maintainer guidance.

Participants and scenarios

Use at least three independent participants: a non-technical reader, an agent/application developer and a prospective node operator. Test first-minute comprehension, the quickest relevant path, mobile/keyboard accessibility and recovery from one realistic failure.

Evidence

Record only consented, content-minimized findings. Convert concrete website defects into private website issues and protocol/documentation ambiguity into public IICP issues.

Acceptance

Publish an anonymized findings summary, prioritized corrections and a repeatable validation script. Do not treat self-review as independent evidence.

Next milestone

Run the published script with exactly the three required participant roles. Accept only consented, anonymized records. Route website defects to the private website tracker and protocol ambiguities here; do not count maintainer self-review as a participant result.

[RobLe3]

External participation lanes are now consolidated in docs/external-evidence-participation.md. This prepares the fixed starting artifact, evidence class, privacy boundary and completion record for this issue without manufacturing an external result or changing any runtime/default. PR #146 passed both hosted gates and merged as f27cca0.

[RobLe3]

The repeatable newcomer session script, blank consent-minimized record, validator, and claim boundaries are now available from docs/external-evidence-participation.md and evidence/newcomer-validation-record-v1.json (merged in PR #146).

No participant result has been manufactured. The next step is execution by three first-time participants, one per required role, followed by the anonymized aggregate summary and prioritized corrections required by this issue. Keep this issue open until those external sessions exist.

[RobLe3]

External ecosystem assessment disposition (2026-08-21): the reviewer could not independently extract current live node/operational facts from the JavaScript-rendered pages, even though /.well-known/iicp-evidence.json, the signed deployment record and JSON stats endpoint exist and currently pass the 10-request ordinary-client verifier.

Add this to newcomer validation: a reviewer starting from the first-level stats/docs pages should discover the raw evidence endpoints, understand which facts are live-only, and distinguish published, deployed and adopted versions without project-private guidance. This is a discoverability/comprehension test, not a reason to duplicate evidence contracts or weaken edge protections. Three-axis projection work is tracked in #188.

[RobLe3]

Newcomer evidence invitation

The public participation packet is ready for one non-technical reader, one developer and one prospective node operator. Each participant follows a short, content-minimized task and publishes only an anonymized result—no prompts, credentials, addresses or full transcript.

Use the session script and evidence template. The useful output is the first blocking phrase or control and the participant's next safe action, not a general opinion score.

[RobLe3]

External-evidence readiness check — 2026-08-24

The canonical campaign index, newcomer blank record and public evidence-access path were revalidated from the current main branch:

• check_external_participation_campaign.py: 5 lanes valid and still awaiting-participant;
• check_newcomer_validation_record.py: blank template valid;
• check_public_evidence_access.py --live: 7 artifacts and 10 live requests valid.

The intake path is complete. No project-authored participant result has been added. This issue remains open for the three consented first-time roles and their anonymized aggregate findings; any reproducible defect should be routed only after that evidence identifies its owner and boundary.

[RobLe3]

The pre-1.0 qualification program now treats newcomer comprehension and failure recovery as a hard evidence dependency rather than optional polish. Protocol/specification reconciliation is tracked in #236 and the cross-component representative workflow program in RobLe3/iicp.network#1107. Project self-tests do not count; failures and abandonments remain in the representative denominator. This issue retains the independent adoption/accessibility scope.