v1.4.0 — Evidence quality

Released 20 July 2026.

Added

• Event-linked evidence records that preserve negative, delayed, and inconclusive outcomes.
• Threshold-boundary investigations with explicit brackets, blockers, and completion criteria.
• Fail-closed evidence adjudication rules, contradiction assessments, and deterministic release-readiness gates.
• Evidence-intelligence pressure scoring and seven bounded acquisition missions.
• Mission-specific intake, privacy screening, human review, conflict disclosure, and controlled dispositions.
• Non-executable promotion plans with exact impact previews, validation requirements, and rollback scope.
• Six-record public reconstruction corpus for Quickdraw, YOLO, and Pair Extraordinaire.
• GitHub-owned live achievement-fragment corpus with exact criteria, linked history, and normalized fingerprints.
• Complete first-party milestone histories for Pull Shark, Galaxy Brain, and Starstruck.
• Weekly live-fragment drift verification.
• Static API discovery expanded to forty public JSON files.

Changed

• Evidence coverage increased from 44.2 at the v1.3.0 baseline to 91.2/100 — Robust.
• Twelve of thirteen canonical claims are now official or confirmed.
• Only Pair Extraordinaire numerical tier thresholds remain below confirmed.
• Quickdraw’s five-minute timing contract is now official.
• Pull Shark milestones are now officially documented as 2, 16, 128, and 1,024 merged pull requests.
• Galaxy Brain milestones are now officially documented as 2, 8, 16, and 32 accepted answers.
• Starstruck milestones are now officially documented as 16, 128, 512, and 4,096 repository stars.
• YOLO’s broad contract is now officially documented as Merged without a review.
• Pair Extraordinaire’s broad contract is now officially documented as coauthored commits on merged pull requests.
• Open contradictions reduced from six to three.

Corrected

• Replaced community-only milestone classifications where GitHub’s live product fragments publish the exact contracts.
• Separated broad trigger certainty from unresolved edge cases, processing behaviour, transfer semantics, and attribution rewrites.
• Removed stale blocked-release wording after the merged-main evidence and operational gates passed.

Included work

• PR #232 — Phases 52–56: evidence quality programme
• PR #233 — Evidence intelligence dashboard
• PR #234 — Targeted evidence acquisition missions
• PR #235 — Mission execution intake and triage
• PR #236 — Mission packet review queue
• PR #237 — Mission evidence promotion planner
• PR #238 — Independent public reconstruction evidence
• PR #239 — GitHub-owned achievement contracts

Verification

Validated from merged main on 20 July 2026:

Gate	Verified	Required	Result
Evidence coverage	91.2	≥ 70.0	PASS
Official or confirmed claims	12	≥ 8	PASS
Claims below confirmed	1	≤ 5	PASS
Open contradictions	3	≤ 3	PASS
Operational health	100/100	100	PASS

• All permanent evidence, content, API, browser, visual, release, and repository-audit workflows passed.
• GitHub-owned fragment criteria and normalized fingerprints were re-fetched and verified live.
• Markdown, links, catalogue consistency, JSON contracts, Jekyll output, search routing, and visual baselines passed.
• The release is published from the merged release-note commit by the repository’s verified immutable release workflow.

Known limitations

• Pair Extraordinaire numerical tier thresholds remain community-reported because GitHub’s x4 detail fragment does not publish the counts.
• YOLO alternate review states and structured merger identity remain unresolved edge cases.
• Pair Extraordinaire squash, rebase, email-mismatch, and rewritten-history behaviour remains unresolved.
• Starstruck repository transfer, organisation ownership, forks, archival, falling counts, and persistence remain unresolved.
• Achievement processing and profile visibility may still be delayed independently of a qualifying event.
• Public product fragments can change; the weekly monitor fails closed on criterion or fingerprint drift.