docs(research): actually ship the 14->15 count provenance - #129
Merged
Conversation
A previous commit intended to add the count_provenance reconciliation but
pushed the pre-reconciliation file (the write and the push used different
paths). This commit ships the corrected graded_reference_set.{json,md},
which record that the Citations track verified 14 sources (8 confirmed) and
the founding Agent-as-a-Judge paper (arXiv:2410.10934) was the +1 added on
that track's own recommendation, giving canonical 15/9/6/0 and 47 entries.
Adds a new section 14 documenting what happened when three further bodies of work were audited against Theorem D: forgekit, hikmah-stack (Rust cognitive kernel v3.0.0), and the wisdom-lens plugin with its two source manuscripts. Findings that changed the paper: - The law is violated inside the repo that states it most explicitly: one host adapter of hikmah-stack binds the completion gate deterministically, the other probabilistically, and the repo's own validator returns ok on the asymmetry because it checks packaging rather than behaviour. Theorem D's predicted failure shape, found in production. - Five kernel components have NO counterpart in this theory: goal-directed symbolic planning, prospective memory with deadlines, privacy-classed persistence, and (a genuine divergence, not an extension) append-only provenance versus our bounded-snapshot handoff. - A correction to our own evidence discipline: hikmah-stack's EVIDENCE.md carries per-document verification dates and a five-rule maintenance policy, and already had a 2026 scoping caveat our own appendix lacked. Practice adopted, not merely noted. Also adds the audit deliverables under audits/ (17 statistics re-verified against fetched primary sources, 0 unverifiable; 17-component crosswalk built after compiling and running the kernel) and Figure 8.
9 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
A previous commit intended to add the count_provenance reconciliation but pushed the pre-reconciliation file (the write and the push used different paths). This commit ships the corrected graded_reference_set.{json,md}, which record that the Citations track verified 14 sources (8 confirmed) and the founding Agent-as-a-Judge paper (arXiv:2410.10934) was the +1 added on that track's own recommendation, giving canonical 15/9/6/0 and 47 entries.
What & why
Checklist
npm testpasses (Node 18/20/22)npm run checkpasses (Biome lint + format)feat:/fix:/docs:…)CHANGELOG.mdupdated under## [Unreleased]forge substrate,forge impact, router/gate, or MCP substrate toolsRisk & rollback
Extra checks (tick if applicable)
npm run typecheckpasses