Skip to content

Accept any recorded or inferred ecosystem id and any mutation tool value #15710

Open
3 of 4 issues completed
Open

Description

@kyle-sexton

Parent

Refs #15703
Source: PLAN Phase 8, topic ecosystem-specifics

Depends on

Outcome

/toolchain:check, /toolchain:lint and /verification:confirm accept any ecosystem id that has a record or was inferred this run, and an unlisted mutation tool value is accepted and reported as having no shipped adapter. Closed lists stop deciding which languages and tools a consumer may use.

Demo: a mutation-testing config with tool: cargo-mutants validates against the schema, and a /testing:audit scan of this repository shows zero changed findings before and after.

Pre-flight first: list every reader of the mutation-testing tool key and of the ecosystem id list across scripts and schemas.

What changes:

  • The mutation-testing schema's tool enum becomes a string with manual reserved, and the config template follows. An unlisted tool keeps the layer, is reported as having no shipped adapter, and is refused for an in-tree run with the manual offer (the write-regime rows stay).
  • Check, lint and confirm accept any covered id; confirm's list today lacks markdown and go. Covered means recorded, or inferred this run with its globs logged.
  • The code-metrics config resolver: a lane id matching a shipped collector selects it; an unknown id is accepted and reported as having no shipped collector.
  • The dead-code shape classifier also reads consumer record globs, so a listed language's sources are nolane, not other.
  • The bash-harness test adapter: this repository's own suite-runner pattern moves to this repository's testing convention file under extend; the scanner test supplies it through an extend fixture, with a negative test for the shipped adapter alone.
  • Each executed adapter touched here names a file or value it cannot read. /architecture:map-containers names the projects it cannot classify.

Method:

  • Tests first in each script's existing suite, with a negative-path test (an unlisted language, an unknown tool value).
  • Skill-text changes ship on a two-version compare.
  • Run budget: 850 runs (18 toolchain check and lint cases, 43 confirm cases and the mutation-testing audit cases, counted first, on the opus and sonnet aliases, three runs per side, after two measuring runs). Select cases with --case, use --runs 3 and a --max-cost-usd ceiling, and start no batch after a usage-limit error.
  • Eval batches run in the one local eval-runner session in the main checkout. Open the pull request with the do-not-merge label; the compare record releases it.

Decisions

  • D2 of the Brief (interview Q2): "no shipped language specifics" covers model-read guidance only; executed adapters, single-tool plugins and security floors stay.
  • Q26: compares run on the opus and sonnet aliases, three runs per side.

Key interfaces

  • "Executed adapters, single-tool plugins and security floors stay; an adapter names what it cannot read instead of skipping it."
  • The record's format is the existing ecosystem-commands schema.

Done when

  • AC4 (id clause): check, lint and confirm accept any ecosystem id that is recorded or inferred this run.
  • AC8: each executed adapter touched names a file or value it cannot read, and an unlisted mutation tool value is accepted and reported as having no shipped adapter (a schema test that tool: cargo-mutants validates passes).
  • The suite-runner pattern is gone from the shipped bash-harness adapter and present once in this repository's testing convention file.
  • A full /testing:audit scan of this repository shows zero changed findings before and after (dogfood).
  • Each touched script's test file passes; the two-version compare is non-inferior and its record is in the pull request body.
  • AC7: nothing is cut here, so no control result is owed.
  • AC10: no new enforcement-placing skill is added here.

Out of scope

  • The unattended flag and the unresolved cell (a later item).
  • Deleting toolchain's bundled files.

Activity

  1. added
    agent-readyFully specified and briefed; eligible for autonomous pickup from the frontier.
    on Nov 12, 2026
  2. github-actions commented on Nov 13, 2026

    @github-actions
    Contributor

    The triage sweep dispatched intake triage for this issue (dispatch 1; 2 runs that reach intake, or 6 dispatches, hand it to a maintainer).

  3. github-actions commented on Nov 13, 2026

    @github-actions
    Contributor

    Size: Large
    Complexity: Medium
    Uncertainty: Known
    Flags: none
    Blocked by: #15705

    Understood: make /toolchain:check, /toolchain:lint and /verification:confirm accept any recorded or inferred ecosystem id. Make the mutation-testing tool value an open string with manual reserved, so an unlisted tool is reported as having no shipped adapter. Also update the code-metrics resolver, the dead-code classifier, the bash-harness adapter's suite-runner pattern, and the "cannot read" naming in executed adapters.

    The issue's outcome and done-when items are exact, so nothing is missing. It depends on open #15705, which is listed as blocked-by. It touches many scripts and schemas plus eval runs, so it may need splitting. The admission note on the issue says it also needs an open parent issue that a listed human filed. No duplicates or merged fixes found.

    Automated triage. A maintainer reviews it.

  4. melodic-automation-lanes commented on Nov 13, 2026

    @melodic-automation-lanes
    Contributor

    Not admitted to the issue lane yet: sub-issue #16063 is open.

  5. melodic-automation-lanes commented on Nov 14, 2026

    @melodic-automation-lanes
    Contributor

    Split at intake (owner Q93). This issue is Size Large, so its done-when is split into 4 sub-issues, each holding its task-list lines as written.

    Automated split by the lanes App. Admission waits on this issue while a sub-issue is open; work-unblock closes it as completed once every sub-issue is closed as completed by a merge or a listed human and its done-when is covered.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    agent-readyFully specified and briefed; eligible for autonomous pickup from the frontier.

    Fields

    Work status

    ready

    Priority

    medium

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions