Repository navigation
Description
Parent
Refs #15703
Source: #16029, part 2 of #15729. The list recorded in #16602 gives /harness-memory:audit ("fix action") the route "call /corrections:codify, presence-gated, with the strongest-first fallback".
Depends on
- Cut the model-read ecosystem specifics from harness-memory #16775, through its open pull request refactor(harness-memory): cut model-read ecosystem examples from audit #16985 (held with
do-not-merge). That pull request edits the C8 bullet list this item rewrites (it drops the examples from "A compiler setting"), so this item starts after refactor(harness-memory): cut model-read ecosystem examples from audit #16985 merges or closes and rebases onto it.
Outcome
C8 ("Enforcement Hierarchy") today lists six mechanisms a CLAUDE.md instruction "could be" and asks the audit to name "which enforcement level it could move to"; the fix action then has no C8 pattern and chooses freely. After this change a C8 instruction that could move up is handed to /corrections:codify when it is available, with the shared strongest-first fallback when it is not, and the fix action gets a C8 pattern that does the same. C8 items 4 and 5 (prose a located check already enforces, and the hold for protected classes) do not change.
Demo: with a corrections stand-in plugin loaded, /harness-memory:audit then fix on a CLAUDE.md holding an import-order rule no check covers reports C8 WARN naming /corrections:codify for the move, and the fix proposal hands the rule to codify instead of proposing a linter rule or a hook itself.
Where the text changes
Line numbers as read on main at a483a2c1; they may move, and #16985 changes the bullet at :301.
| File and lines | Present wording | Change |
|---|---|---|
plugins/harness-memory/skills/audit/reference/criteria.md:294-308 (C8 "How to check", items 1-3) |
"For each rule/instruction, check if it could be: A compiler setting (nullable, warnings-as-errors); A static-analyzer or linter rule; An architecture test; A pre-commit/pre-push git hook; A CI gate; A Claude Code hook (deterministic) ... Include which enforcement level it could move to" | Item 1 asks whether a check could decide the instruction, without a mechanism list; item 3 hands the level to /corrections:codify (gate plus shared fallback). Add a Provenance line that marks the new text as this repository's rule, as items 4 and 5 already carry, so the update action does not overwrite it |
same file, :127-128 (C3 placement table) |
"Deterministic enforcement | Hooks (guaranteed execution)"; "Compile-time/build-time rules | Analyzers, linters, architecture tests" | One row: content a check could decide goes to /corrections:codify for its mechanism, with the fallback |
plugins/harness-memory/skills/audit/context/fix.md ("Common fix patterns", no C8 section today) |
none | Add "C8 (Enforcement Hierarchy) fixes": hand each WARN that could move up to /corrections:codify rules <the file> or /corrections:codify "<the rule>"; the prose stays until the check exists, then goes in that change (C8 item 4) |
Decisions
-
Design Q21 (Resolve ecosystem specifics from the consumer repository and align the enforcement ladder with the talk's pyramid #15703): a skill that carries no ladder copy reads the ladder by calling codify; only a skill that applies its own selection rule holds a synced copy and a Selection rules row.
-
docs(enforcement-ladder): add the architecture and retro Selection rules rows in ladder 3.2.0 #16602 recorded this skill under "call
/corrections:codify, presence-gated, with the strongest-first fallback (12; change, filed as follow-on items by part 2, List enforcement-placing skills and give each one route to the ladder - part 2 of 2 #16029)". -
The fallback wording is shared by all nine follow-ons of List enforcement-placing skills and give each one route to the ladder - part 2 of 2 #16029. After the gate sentence ("when
/corrections:codifyis among the available skills, hand it the choice of mechanism; it picks the strongest rung of the enforcement ladder that can decide the rule and proves the check"), the fallback is exactly:Otherwise, propose the strongest mechanism that can decide the rule: a change that makes the mistake impossible first, then a check that blocks it, then a test, and an instruction only when nothing a machine runs can decide it. Say that
/corrections:codifywas not available, so the proposal is unproven.
Key interfaces
- "A skill that carries no ladder copy calls
/corrections:codify, presence-gated, with a two-sentence strongest-first fallback. Only a skill that applies its own selection rule holds a synced copy and gets a row in the ladder's Selection rules table." (List enforcement-placing skills and give each one route to the ladder #15729) - The seam-phrasing convention: an availability gate, a stated fallback beside it, the plugin named and never the marketplace.
- The instruction exception register keeps owning which instructions hold their prose (C8 item 5).
Done when
- C8 items 1-3 hand the move up the hierarchy to
/corrections:codifybehind the availability gate, followed by the shared fallback, and list no mechanisms of their own; items 4 and 5 are unchanged. - The new C8 text carries a Provenance line so the
updateaction does not overwrite it. -
context/fix.mdhas a C8 pattern that hands each instruction to/corrections:codifyand keeps the prose until the check exists. - The skill gains no ladder copy or Selection rules row, and no rung list is restated.
- The pull request body quotes the sentence that picked a mechanism from a private list before the change. If the recheck finds none, the skill is left unchanged and the finding is reported on List enforcement-placing skills and give each one route to the ladder - part 2 of 2 #16029.
- One new eval case loads a corrections stand-in plugin, in the shape of session-flow's
re-mechanical-finding-check-proposalcase, and grades that a C8 instruction no check covers is handed to/corrections:codifyand the audit picks no mechanism itself. - The pull request body names the existing case whose run reads the changed text (candidates:
audit-c8-enforced-prose,audit-fix-needs-prior-audit), or says none does. The lane runs no eval. - A changelog fragment under
.changes/harness-memory/describes the change. - AC10: no new enforcement-placing skill is added here.
- Open the pull request with
Closesthis issue in its body, then stop: the lane adds no label.
After the pull request opens (host session, not the lane)
The pull request stays held with do-not-merge for the eval; a host session runs only the eval and posts its record; the merge lane releases the hold when the record matches the head.
- Hold line: the hub's sweep (or a host session) adds
do-not-mergeand posts, from a listed human's account, a comment with the line<!-- pr-hold reason=eval -->on a line of its own. - Run budget. List enforcement-placing skills and give each one route to the ladder #15729 asks for "that skill's cases on the
opusandsonnetaliases, three runs per side, and no more than 240 runs per plugin". The owner decision of 2026-11-16 in the owner-decisions record (hop 4 Q15, which names List enforcement-placing skills and give each one route to the ladder #15729, amended by hop 7 Q3) replaces the run counts: per changed skill, one targeted before-and-after case onsonnetat 3 runs a side;opusruns only to confirm a sonnet loss; a new case gets a 1 to 3 run smoke first. Select cases with--case, use--runs 3and a--max-cost-usdceiling, and start no batch after a usage-limit error. Ceiling for this plugin: 1 skill x (3 smoke + 6 sonnet + 6 opus confirm) = 15 runs, under the parent's 240. - Control: the removed mechanism list is a partial cut, so the pull request body carries a detection-control result. No paired no-loss compare (#15703 comment 6906800940).
- On a LOSS reading: a trace read of each dropped case and an
opusconfirm; only a loss traced to the changed text counts. - Record: the verdict and the control result go into the pull request body. When the run is VALID and shows no loss, the host session posts, from a listed human's account, a comment with the line
<!-- eval-pair verdict=no-loss validity=VALID head=<40-hex head SHA> -->naming the current head, keeping "do not merge" anddo-not-mergeout of that comment (docs/contracts/loop-lane/README.md, "Cross-lane PR hold"). A loss or a run that is not VALID leaves the hold in place; say so on the pull request. - Release: pr-merge's scheduled step "Release eval holds whose record matches the head" removes
do-not-mergeonce that record names the current head and no other hold stands. A push after the record needs a new record.
Out of scope
- C8 items 4 and 5, the deterministic spine, and every other check.
- The harness-memory ecosystem cut (Cut the model-read ecosystem specifics from harness-memory #16775 and its pull request refactor(harness-memory): cut model-read ecosystem examples from audit #16985).
Related
- Resolve ecosystem specifics from the consumer repository and align the enforcement ladder with the talk's pyramid #15703 (container)
- List enforcement-placing skills and give each one route to the ladder #15729
- List enforcement-placing skills and give each one route to the ladder - part 2 of 2 #16029
- docs(enforcement-ladder): add the architecture and retro Selection rules rows in ladder 3.2.0 #16602
Metadata
Metadata
Assignees
Labels
Type
Fields
Work status
Priority
Activity
github-actions commentedon Nov 18, 2026
Size: Medium
Complexity: Medium
Uncertainty: Known
Flags: none
Blocked by: none
Blast radius: harness-memory audit skill text (criteria.md C8 and C3 row, context/fix.md), one new eval case and a changelog fragment; read by the audit skill and anyone running
/harness-memory:audit.Reversible: yes; revert the text, case and fragment.
Open questions: none
Understood: the audit's C8 items 1-3, the C3 placement row and fix.md should hand a movable instruction to
/corrections:codifybehind an availability gate with the shared fallback, with no mechanism list or ladder copy, plus a Provenance line and a changelog fragment. The two eval/recheck items are person-run checks, already split out by the refinement comment. The dependency #16985 is a pull request (for #16775), so it is not listed as a blocking issue; start after it merges or closes. Related sibling hand-offs: #17158, #17163, #17159, #17175. #16928 may touch the same files. No duplicate or merged fix found.Automated triage. A maintainer reviews it.
melodic-automation-lanes commentedon Nov 18, 2026
Not admitted to the issue lane: its done-when asks for what an unattended lane run cannot perform: item 5 (tracker-output: "The pull request body quotes the sentence that picked a mechanism from a priv..."), item 6 (eval: "One new eval case loads a corrections stand-in plugin, in the shape of sessio..."); recommended split: keep the other items here as the admissible issue and move these to a child issue for the host queue (
host-run). Admission takes the issue once each of those is met, when intake triages it again (.github/actions/resolve-config/admission-rules.json).melodic-automation-lanes commentedon Nov 18, 2026
Refinement. A read-only run explored this repository, searched its issues and pull requests, and read the allowed docs to draft what follows. On an issue a listed human filed, it counts as the issue's spec once posted, while unedited; it is still model text, so check it.
Intended outcome: The harness-memory audit's C8 items 1-3, its C3 placement row and fix.md hand a movable instruction to /corrections:codify behind an availability gate with the shared fallback, with no mechanism list or ladder copy, plus a Provenance line and a changelog fragment.
Split. Done-when items 5, 6 of the issue body need a person: they are checks a person runs, listed below, and the lane does not do them. This issue keeps the rest of its done-when.
Not done by the lane: checks a person runs
External state: none
Dependencies
Related work
Sources read
Automated refinement by the lanes App. Intake triages the issue again with it.
melodic-automation-lanes commentedon Nov 18, 2026
{"v":1,"actor_id":337595936,"title_sha256":"722e70f06f08839b0d96b7fd3300fccb58100c532b8f8e1ed55fd79263509de6","body_sha256":"d1737855c9c013b39e6f643beb0576444965f23417cf2337fe9ab1108113adab"}
melodic-automation-lanes commentedon Nov 18, 2026
{"v":1,"actor_id":337595936,"title_sha256":"722e70f06f08839b0d96b7fd3300fccb58100c532b8f8e1ed55fd79263509de6","body_sha256":"d1737855c9c013b39e6f643beb0576444965f23417cf2337fe9ab1108113adab"}
github-actions commentedon Nov 18, 2026