Repository navigation
Conversation
… claims Adds the owner's preservation-first policy (2026-11-04): main is the authoritative research record. Positive, negative and inconclusive results, failed approaches and their supporting files may stay on main with their provenance, assumptions, verification status and limitations stated. A claim is presented as established only when its evidence meets the verification standard for its kind. Retention promotes nothing. A record's status is stated as independently verified, partially verified, exploratory or known invalid. Unsupported conclusions are corrected where stated while their evidence stays. Removal needs a specific justification, a dependency audit and separate owner authorization. Independent of #1097, which edits §A.42; this adds a new section at the end of the file. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013T6pfKuv5mi3MXtKjH2tZb
- F1: §A.42 item 5 and §A.28 limit what is added to main and are not a ground for removing research evidence already there. - F2: a frozen record's unsupported conclusion is corrected by a new native round (§A.42 item 6). - F3: annotation over deletion applies to research evidence; a manuscript claim that loses its support is handled as §A.27 and §A.30 require. - F4: a record distinguishes the four kinds of result, in the record or, for a frozen record, in a mutable record on main that cites it. - F5: the owner's "provided that" condition is restored in the opening. - F6: item 2 applies to results that do not meet item 1's standard, so a verified negative result can be established (§A.29). - F7, F8, F9: separator line; a resolvable cross-reference; the adoption line narrowed to what is recorded. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013T6pfKuv5mi3MXtKjH2tZb
Owner
Author
|
Closing without merging, combined into #1097 (owner request, amaybaum, Claude Code session chat, 2026-11-04: "Why do we need 2 separate open PR's for AGENTS.md file changes?", then "Yes please do").
Generated by Claude Code |
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
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.
Owner decision (amaybaum, Claude Code session chat, 2026-11-04):
The owner's recommended policy:
Change. One file,
AGENTS.md. It adds a new section at the end, §A.46 "Research records onmainand established claims", in two commits:3a8920a8adds the section;c97de3ddapplies the independent review.What the section says:
mainis the authoritative research record. Research evidence may stay whether or not its programme is complete, provided its provenance, assumptions, verification status and limitations are stated accurately. §A.42 item 5 and §A.28 limit what is added tomain; they are no ground for removing what is already there.mainthat cites it.For the owner to confirm (review finding F6). The owner's text says "Incomplete, negative and inconclusive results must not be represented as established". Read literally, that would stop a verified negative result from ever counting as established, which conflicts with item 1 and with §A.29 ("A route closed by controlled computation is progress"). Item 2 therefore says "… that do not meet item 1's standard". The owner can restore the literal wording if preferred.
Review.
3a8920a8, returned CHANGES REQUIRED with two blocking findings:c97de3ddapplies F1–F3 and nits F4–F9.c97de3ddcomes before the merge request.Related.
Checks.
Gates.
🤖 Generated with Claude Code
https://claude.ai/code/session_013T6pfKuv5mi3MXtKjH2tZb