agentleFS
tikalktikalk/adlc-team-skills/skills/change/change-clarify/SKILL.md

change-clarify

Skill145 starsChanged 21 days agoMITView on GitHub

Use when reviewing, accepting, rejecting, or deferring ChDRs mined by change-init, validating inferred decisions against their git and issue evidence before promotion to project memory.

The file

Show its rules and text

We split this skill into the rules it gives an agent

29 rules. About 0 are common, 19 are distinctive, and 10 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈19

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 10

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check inferred-rationale confidence and flag low-confidence or unsupported claims

    line 16
  2. Distinctive:

    Ensure Consequences (reverts/fix chains) are recorded where observed

    line 18
  3. Very specific:

    Update ChDR statuses in {REPO_ROOT}/.adlc/drafts/chdr/ChDR-{NNN}.md

    line 19
  4. Distinctive:

    No pending ChDRs: if no ChDRs have status Discovered/Proposed, there is nothing to clarify

    line 31
  5. Very specific:

    Reviewing CDRs: use /team-levelup (CDRs and ChDRs are distinct record types with separate gates)

    line 32
  6. Distinctive:

    Direct editing: do not use this skill to bypass the review workflow

    line 33
  7. Very specific:

    Read all {REPO_ROOT}/.adlc/drafts/chdr/ChDR-*.md files and filter:

    line 97
  8. Distinctive:

    If no pending ChDRs:

    line 108
  9. Very specific:

    Check if the draft being reviewed uses the lightweight draft template (indicated by presence of type, evidence, source, revisit-when fields …

    line 117
  10. Distinctive:

    Read the draft's captured fields (Context, Decision, Rejected Alternatives, Reason)

    line 123
  11. Distinctive:

    If already full format, proceed with normal review.

    line 133
  12. Very specific:

    Provenance: every non-trivial sentence in ### Decision references a SHA (\b[0-9a-f]{7,40}\b) or URL (https?://)

    line 143
  13. Very specific:

    Write every clarification in Simplified Technical English (short sentences, one idea each, active voice, imperative asks, named referents, …

    line 180
  14. Distinctive:

    Ask more than one question at a time

    line 187
  15. Distinctive:

    Session limit: Default 5 ChDRs per session.

    line 189
  16. Distinctive:

    Wait for user input before proceeding.

    line 226
  17. Very specific:

    Update status to ### Status: Accepted.

    line 230
  18. Very specific:

    Update status to ### Status: Rejected with reason.

    line 261
  19. Distinctive:

    Keep status as-is. Add note:

    line 265
  20. Distinctive:

    After EACH ChDR interaction, immediately update the file.

    line 277
  21. Distinctive:

    Ask exactly ONE question per response

    line 323
  22. Distinctive:

    Wait for user input before proceeding

    line 324
  23. Distinctive:

    Update ChDR file after each decision

    line 328
  24. Distinctive:

    Never accept or reject without explicit user choice

    line 333
  25. Distinctive:

    Do not assume user preference

    line 334
  26. Very specific:

    If any ChDRs were Accepted, handoff to /change-publish:

    line 345
  27. Very specific:

    After accepting ChDRs, run /change-publish to promote them to project memory and regenerate the boot-facing chdr.md index.

    line 370
  28. Distinctive:

    All reviewed ChDR files updated with new status and clarification metadata.

    line 374
  29. Distinctive:

    No ChDRs were auto-accepted or auto-rejected without user input.

    line 377

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

change-clarify

What this skill does

Review pending ChDRs (status Discovered or Proposed) and decide their fate: Accepted, Rejected, or Deferred.

This is the quality gate for Change Decision Records — the human checkpoint before inferred-from-history decisions are promoted to project memory and injected into future sessions by team-boot:

  • Validate that inferred decisions are supported by cited commit SHAs / issue URLs (provenance — the context-poisoning circuit breaker)
  • Check inferred-rationale confidence and flag low-confidence or unsupported claims
  • Confirm issue-link validity and that fetched issue summaries are accurate (not pasted verbatim)
  • Ensure Consequences (reverts/fix chains) are recorded where observed
  • Update ChDR statuses in {REPO_ROOT}/.adlc/drafts/chdr/ChDR-{NNN}.md
  • Regenerate {REPO_ROOT}/.adlc/drafts/chdr/chdr.md index

This is an interactive command. Present exactly one ChDR per interaction and wait for user input.

When to use

  • After /change-init: validate history-mined ChDRs before promotion
  • Periodic review: clean up stale pending ChDRs

When NOT to use

  • No pending ChDRs: if no ChDRs have status Discovered/Proposed, there is nothing to clarify
  • Reviewing CDRs: use /team-levelup (CDRs and ChDRs are distinct record types with separate gates)
  • Direct editing: do not use this skill to bypass the review workflow

Process

User Input

$ARGUMENTS

You MUST consider the user input before proceeding (if not empty).

Examples of User Input:

  • "ChDR-001 ChDR-003" — focus on specific ChDRs
  • "all" — clarify all pending ChDRs
  • Empty input: clarify all ChDRs with status "Discovered" or "Proposed"

Flags

  • --all: clarify all pending ChDRs (same as empty input)
  • --limit N: limit to N clarifications per session (default: 5)

Role & Context

You are acting as a Change Validator reviewing inferred-from-history decisions. Unlike CDR clarification (which validates team-wide patterns), your focus is evidence fidelity:

  • Was the Decision actually inferable from the cited commits, or did the agent invent it?
  • Is the confidence level honest (a HIGH-confidence claim with a terse commit + no issue link is suspicious)?
  • Are issue summaries accurate and not leaking sensitive detail verbatim?
  • Are reverts/fix chains recorded where they exist (the highest-value content)?

ChDR Quality Checklist

Each ChDR should have:

  • ### Context explaining why the change happened (issue summary or commit messages)
  • ### Decision with inferred decision + confidence level
  • Every Decision claim cites a SHA or issue URL (provenance — non-negotiable)
  • ### Consequences (reverts/fix chains, or explicit "None observed")
  • ### Evidence with concrete SHAs, files, and issue links
  • Issue summaries are paraphrased, not raw paste (security)
  • Status is accurate

Outline

  1. Load Pending ChDRs (Phase 1): parse ChDR files with status Discovered/Proposed
  2. Pre-Validation (Phase 2): skip ChDRs missing required sections or with unprovenanced Decision claims
  3. Gap Identification (Phase 3): list clarification needs
  4. Sequential Clarification (Phase 4): one ChDR per interaction
  5. Update ChDRs (Phase 5): write status and clarification metadata after each decision
  6. Regenerate Index (Phase 6): update chdr.md
  7. Summary (Phase 7): present results

Execution Steps

Phase 1: Load Pending ChDRs

Run setup script:

scripts/bash/setup-change-clarify.sh

Read all {REPO_ROOT}/.adlc/drafts/chdr/ChDR-*.md files and filter:

  • Include: ### Status: **Discovered** or ### Status: **Proposed**
  • Skip: ### Status: **Accepted**, **Rejected**, **Published**, **Deprecated**

If the setup script is unavailable or fails, resolve manually:

  1. REPO_ROOT — walk up from cwd to find .adlc/, or git rev-parse --show-toplevel.
  2. CHDR_DRAFTS_DIR — REPO_ROOT/.adlc/drafts/chdr
  3. PENDING_COUNT — count files with Discovered/Proposed status.

If no pending ChDRs:

No pending ChDRs found.
Run /change-init to mine git history first.

Step 0: Detect Lightweight Draft Format

Check if the draft being reviewed uses the lightweight draft template (indicated by presence of type, evidence, source, revisit-when fields in frontmatter and ## Rejected Alternatives / ## Reason body sections without the full formal template sections).

If lightweight:

  1. Read the draft's captured fields (Context, Decision, Rejected Alternatives, Reason)
  2. Transform to full formal template:
    • ADR: full MADR format with Decision Drivers, Considered Options, Pros/Cons, Constitution Alignment, Related ADRs
    • PDR: full PDR format with Market Forces, Consequences, Alternatives Considered, Links
    • ChDR: full ChDR format with Issue Links, Commits, Consequences, Evidence
    • CDR: full CDR format with Context Type, Target Module, Descriptor, Evidence
    • EVAL: full eval format with Error Analysis, Pass/Fail Examples, Implementation Notes
  3. Enrich from session context (add details the lightweight draft may have omitted)
  4. Present the enriched draft for review

If already full format, proceed with normal review.

Phase 2: Pre-Validation

For each pending ChDR, check required sections:

  • ### Context
  • ### Decision (with at least one confidence marker)
  • ### Consequences
  • ### Evidence
  • Provenance: every non-trivial sentence in ### Decision references a SHA (\b[0-9a-f]{7,40}\b) or URL (https?://)

Skip invalid ChDRs and report:

## Skipped ChDRs

| ChDR | Issue | Action |
|---|---|---|
| ChDR-XXX | Decision claims lack SHA/URL provenance | Re-run /change-init or add evidence manually |

Phase 3: Gap Identification

Generate a gap report:

## ChDR Clarification Report

| ChDR | Title | Gap Type | Severity |
|---|---|---|---|
| ChDR-001 | [Title] | Low-confidence decision | MEDIUM |
| ChDR-002 | [Title] | Issue summary too verbose | LOW |

Gap types:

  • Low-confidence decision: Decision marked LOW confidence — consider rejecting or deferring
  • Unprovenanced claim: a Decision sentence without SHA/URL (should have been caught in Phase 2)
  • Issue summary too verbose: pasted verbatim instead of paraphrased (security)
  • Missing consequences: Consequences section empty but reverts observed in window
  • Stale: cluster's commits later reverted (the decision was reversed — mark for rejection or status reversal)

Phase 4: Sequential Clarification

CRITICAL: Present exactly ONE ChDR per interaction.

Write every clarification in Simplified Technical English (short sentences, one idea each, active voice, imperative asks, named referents, plain words). Read the contract at skills/authoring/references/ste-rules.md before drafting questions.

Count crossed boundaries (record, lane, team, now) for each gap: 0–1 boundaries use the plain STE format below, however severe the gap; 2 or more boundaries draw the full ASCII decision map from the reference (WHAT/WHY, crossed boundaries, reach fan-out, candidates × discriminating axes, IF-UNDECIDED). Render unknown gaps as ? slots, never prose complaints. Do NOT:

  • Present multiple ChDRs together
  • Auto-select actions
  • Proceed without explicit user input
  • Ask more than one question at a time

Session limit: Default 5 ChDRs per session. User can say "done" to exit early.

For each ChDR:

## ChDR-{ID}: {Title}

**Status**: {status}
**Issue Links**: {links or "none"}
**Commits**: {sha list}

### Current Content

**Context**:
{context}

**Decision** (confidence: {level}):
{decision}

**Consequences**:
{consequences}

**Evidence**:
{evidence}

### Choose Action

| Option | Action |
|---|---|
| A | **Accept** — Approve for promotion to memory |
| B | **Reject** — Decline with reason |
| C | **Defer** — Skip for now, keep pending |
| D | **Accept all remaining** — Accept this ChDR and all pending ChDRs without further review |

Reply with your choice (A/B/C/D).

Wait for user input before proceeding.

Action A: Accept

Update status to ### Status: **Accepted**. Add clarification metadata:

### Clarification

- **Date**: [YYYY-MM-DD]
- **Action**: Accepted
- **Rationale**: [summary of review]

Action D: Accept All Remaining

Accept the current ChDR (Action A), then iterate remaining pending ChDRs marking each ### Status: **Accepted** with bulk metadata. Skip per-ChDR presentation. Proceed to Phase 6.

Action B: Reject

Ask for reason:

### Decision: Reject

| Option | Reason |
|---|---|
| A | Decision was reversed (later reverted) |
| B | Inferred rationale unsupported by evidence |
| C | Project-specific / not worth recording |
| D | Duplicate of an existing ChDR or ADR |

Reply with your choice.

Update status to ### Status: **Rejected** with reason.

Action C: Defer

Keep status as-is. Add note:

### Clarification

- **Date**: [YYYY-MM-DD]
- **Action**: Deferred
- **Reason**: [need more context / waiting on team / low priority]

Phase 5: Update ChDR Files

After EACH ChDR interaction, immediately update the file. Do not batch at the end.

Phase 6: Regenerate Index

Regenerate {REPO_ROOT}/.adlc/drafts/chdr/chdr.md by listing all ChDR-*.md files and building the markdown table from single-line fields (### Status:, ### Date:, ### Issue Links:, ### Commits:, ### Descriptor:).

Phase 7: Summary

## Change-Clarify Summary

**ChDRs Reviewed**: N
**Accepted**: N
**Rejected**: N
**Deferred**: N

### Accepted (Ready for Promotion)

| ChDR | Title |
|---|---|
| ChDR-001 | [Title] |

### Rejected

| ChDR | Reason |
|---|---|
| ChDR-003 | Decision was reversed (later reverted) |

### Deferred

| ChDR | Title |
|---|---|
| ChDR-004 | [Title] |

### Next Steps

1. **Accepted**: Run `/change-publish` to promote to `docs/adlc/memory/chdr/`
2. **Deferred**: will appear in next clarify session
3. **Remaining**: run `/change-clarify` again to continue

Key Rules

One-at-a-Time

  • Present exactly ONE ChDR per response
  • Ask exactly ONE question per response
  • Wait for user input before proceeding

Immediate Writes

  • Update ChDR file after each decision
  • Regenerate index after session ends

No Auto-Approval

  • Never accept or reject without explicit user choice
  • Do not assume user preference

Provenance Is the Gate

  • A Decision claim without SHA/URL provenance is a poisoning risk — reject or require evidence before accepting
  • LOW-confidence decisions are acceptable to accept if evidence is solid; HIGH-confidence with no evidence is not

Workflow Guidance & Transitions

After /change-clarify

If any ChDRs were Accepted, handoff to /change-publish:

{
  "command": "clarify",
  "accepted": ["ChDR-001", "ChDR-002"],
  "rejected": ["ChDR-003"],
  "deferred": ["ChDR-004"]
}

Complete Clarify Flow

[Pending ChDRs exist]
    ↓
/change-clarify
    ↓
[One ChDR at a time] → Accept / Reject / Defer
    ↓
[Run /change-publish] → Promote accepted ChDRs to docs/adlc/memory/chdr/

Next Steps

After accepting ChDRs, run /change-publish to promote them to project memory and regenerate the boot-facing chdr.md index.

Verification

  • All reviewed ChDR files updated with new status and clarification metadata.
  • chdr.md drafts index regenerated.
  • Accepted ChDRs are ready for /change-publish.
  • No ChDRs were auto-accepted or auto-rejected without user input.
  • Accepted ChDRs all have provenance on Decision claims.

Context

$ARGUMENTS

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.