docs: restore PTCNA construction and evaluation boundary

[erinepshovel-code]

What changed

• removes the obsolete policy that made upstream evidence and simpler baselines prerequisites for constructing the selected PTCNA architecture
• makes faithful PTCNA construction, dependable fallback construction, and later evaluation three distinct work streams
• keeps Does it work? as the critical falsifiability question and requires its workload, comparator, metrics, thresholds, limits, stopping rules, and failure propagation to be frozen before outcome inspection
• preserves the UCNS authority boundary: local PTCNA prime/ring construction may proceed, while UCNS attribution still requires an exact pinned producer receipt
• replaces the completed Blockers for me section with the live construction/falsifiability contract
• propagates the applicable current skill-lib doctrine and updates the local source pin

Why

The previous repository plan incorrectly converted evidence discipline into a permission gate. Falsifiability governs the verdict; it does not veto construction of the selected inquiry target. The fallback protects useful operation if PTCNA fails and does not replace or redefine PTCNA.

Authority and provenance

• PTCNA base: f0208bf03f36be09d85a369c69d913fa94b71b0e
• skill-lib doctrine: b4234ca29529f56526541df8deb58c2c19570792
• METAPAT consultation: 53315e30c54aba881a5b48cbf395890e83ab05c5, POSTULATES.md, Seventh Postulate
• runtime evidence identity remains 4509d33419aa25e6a9cfef415055e378b8f37edc
• no mathematical, semantic, measurement, empirical, or architectural-validity status transfers

Validation

• remote contents match the intended five-file patch byte-for-byte
• one commit ahead of current main; no unrelated files
• repository-plan JSON parses
• frozen report schema/version/blob identity preserved
• every cross-repository relation retains authority_transfer: false
• prohibited construction-gate language is absent
• faithful construction and the critical Does it work? verdict are both explicit

GitHub Actions supplies the repository-local executable gates.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 6e6baabaf4

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Pin the report to the commit containing its claims

Advancing generated_at while retaining source.commit as 4509d334... leaves this machine-consumed report pinned to a revision that does not even contain repository-plan-report.json (the file was introduced later in 72ad542...), much less the newly rewritten frontier and actions. When the portfolio aggregator uses this source identity for provenance or staleness checks, it cannot reproduce the reported plan and may present the new claims as belonging to the old runtime commit; pin the report to a commit containing these claims, or explicitly separate runtime-evidence and report-source identities.

Useful? React with 👍 / 👎.