Skip to content

feat(work-items): make the AI disclaimer a setting, default off #8650

Description

@kyle-sexton

Refs: #7578
Refs: #7555
Decision: #7555 (comment)

Agent Brief

Type: Feature
Summary: The AI disclaimer becomes a setting, ai_disclaimer, bundled default off, with a per-user userConfig value and a repository value in docs/conventions/work-items.yaml. This repository sets it off.

Problem: Every autonomous lane post starts with the disclaimer, with no way to turn it off. In this repository lane posts already come from a bot account, so the prefix repeats what the author line says.

Current behavior:

  • plugins/work-items/reference/ai-disclaimer.md:3-5: "When an autonomous lane creates a tracker item or comment without a human turn, prefix the body with this disclaimer". There is no setting.
  • Readers that apply it: plugins/work-items/skills/triage/SKILL.md:202, 266-269, triage/context/apply-outcome.md, attend-queue/SKILL.md, work-loop/SKILL.md, work-loop/reference/objection-window.md, reference/dogfood-filing.md:68-77, and docs/conventions/loop-lane/README.md:1165-1168; evals in triage, attend-queue and work-loop assert it.
  • The settings pattern to reuse: triage/SKILL.md:130-169 resolves triage_repro_count from the manifest default, the per-user userConfig value and the repository's docs/conventions/work-items.yaml (schema plugins/work-items/schemas/work-items.schema.json). That file does not exist in this repository yet.

Desired behavior:

  • plugins/work-items/.claude-plugin/plugin.json userConfig gains ai_disclaimer (boolean, default false), and the schema gains the key.
  • ai-disclaimer.md states once: the key, its default, the layers and their order (the same as triage/SKILL.md:140-143), and that when it resolves false no prefix is added. Each reader points there instead of saying "prefix every" unconditionally.
  • New docs/conventions/work-items.yaml with the schema comment line and ai_disclaimer: false.
  • Evals that assert the prefix gain the setting's state in their prompt, plus one case per changed skill where it resolves false and the post carries no prefix.

Key interfaces:

  • Setting ai_disclaimer (boolean, default false; repository and per-user levels).
  • plugins/work-items/reference/ai-disclaimer.md (single statement of the rule).

Done when:

  • The diff adds ai_disclaimer with "default": false to plugins/work-items/.claude-plugin/plugin.json and to plugins/work-items/schemas/work-items.schema.json, and adds docs/conventions/work-items.yaml with ai_disclaimer: false (read from the diff).
  • Every reader listed under Current behavior points at ai-disclaimer.md for the condition (read from the diff).
  • plugins/skill-quality/scripts/check-evals-quality.sh exits 0 on each changed evals.json (triage, attend-queue, work-loop).
  • scripts/check-changelog-fragments.sh --check passes, with one minor fragment for work-items.
  • CI's check-plugins passes on the PR (a PR check; this run does not run it).
  • CI's check-skills passes on the PR (a PR check; this run does not run it).

Out of scope:

  • The wording of the disclaimer itself.
  • Disclaimers outside work-items lanes.

Origin:

  • Disclaimer on every triage post: Matt's (triage SKILL.md:13-17). No disclaimer: hers (Lauren's Slack triage).
  • A setting, default off, repository value and per-user value: the owner's (Q540), confirmed 2026-10-20. This repository sets off because lane posts come from a bot account: ours, confirmed by the owner.

Do not run with: the Phase 11 claim-and-classes slice when filed (it edits triage, work-loop and attend-queue).

Size Medium, Complexity Low, Uncertainty Known.


Slice of #7578 (Phase 11), filed from the owner's 2026-10-20 disposition round.

Activity

  1. added
    status: readyTriaged, unblocked, and fully specified; eligible to pick up.
    on Oct 20, 2026
  2. github-actions commented on Oct 26, 2026

    @github-actions
    Contributor

    The triage sweep dispatched intake triage for this issue (attempt 1 of 2).

  3. github-actions commented on Oct 26, 2026

    @github-actions
    Contributor

    Proposed work class: work-class: scoped
    Basis: automated intake triage; the lanes App applied this class.

    Size: Medium
    Complexity: Low
    Uncertainty: Known
    Flags: none
    Blocked by: none

    Understood: add an ai_disclaimer setting (boolean, default false) in the work-items plugin's userConfig and schema. Add a repository value in a new docs/conventions/work-items.yaml, set to false. State the rule once in ai-disclaimer.md and have each reader point there. Update the evals to match. The brief is exact and the owner's decision is linked, so nothing is missing. I found no likely duplicates. The CI-only done-when items (check-plugins, check-skills) are PR checks.

    Automated triage, applied by the lanes App.

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.status: readyTriaged, unblocked, and fully specified; eligible to pick up.

    Fields

    No fields configured for Feature.

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions