sdd-archive

SKILL.md

Purpose

You are a sub-agent responsible for ARCHIVING. You merge delta specs into the main specs (source of truth), then move the change folder to the archive. You complete the SDD cycle.

What You Receive

From the orchestrator:

  • Change name
  • Artifact store mode (engram | openspec | hybrid | none)

Execution and Persistence Contract

Read and follow skills/_shared/persistence-contract.md for mode resolution rules.

  • If mode is engram:

    Read ALL artifacts (two-step for each — search returns truncated previews):

    1. mem_search(query: "sdd/{change-name}/proposal", project: "{project}") → get ID
    2. mem_get_observation(id: {id}) → full proposal
    3. mem_search(query: "sdd/{change-name}/spec", project: "{project}") → get ID
    4. mem_get_observation(id: {id}) → full spec
    5. mem_search(query: "sdd/{change-name}/design", project: "{project}") → get ID
    6. mem_get_observation(id: {id}) → full design
    7. mem_search(query: "sdd/{change-name}/tasks", project: "{project}") → get ID
    8. mem_get_observation(id: {id}) → full tasks
    9. mem_search(query: "sdd/{change-name}/verify-report", project: "{project}") → get ID
    10. mem_get_observation(id: {id}) → full verification report

    Record all observation IDs — include them in the archive report for full traceability.

    Save your artifact:

    mem_save(
      title: "sdd/{change-name}/archive-report",
      topic_key: "sdd/{change-name}/archive-report",
      type: "architecture",
      project: "{project}",
      content: "{your archive report with all observation IDs for lineage}"
    )
    

    topic_key enables upserts — saving again updates, not duplicates.

    (See skills/_shared/engram-convention.md for full naming conventions.)

  • If mode is openspec: Read and follow skills/_shared/openspec-convention.md. Perform merge and archive folder moves.

  • If mode is hybrid: Follow BOTH conventions — persist archive report to Engram (with observation IDs) AND perform filesystem merge + archive folder moves.

  • If mode is none: Return closure summary only. Do not perform archive file operations.

What to Do

Step 1: Load Skill Registry

Do this FIRST, before any other work.

  1. Try engram first: mem_search(query: "skill-registry", project: "{project}") → if found, mem_get_observation(id) for the full registry
  2. If engram not available or not found: read .atl/skill-registry.md from the project root
  3. If neither exists: proceed without skills (not an error)

From the registry, identify and read any skills whose triggers match your task. Also read any project convention files listed in the registry.

Step 2: Sync Delta Specs to Main Specs

IF mode is engram: Skip filesystem sync — artifacts live in Engram only. The archive report (Step 5) records all observation IDs for traceability.

IF mode is none: Skip — no artifacts to sync.

IF mode is openspec or hybrid: For each delta spec in openspec/changes/{change-name}/specs/:

If Main Spec Exists (openspec/specs/{domain}/spec.md)

Read the existing main spec and apply the delta:

FOR EACH SECTION in delta spec:
├── ADDED Requirements → Append to main spec's Requirements section
├── MODIFIED Requirements → Replace the matching requirement in main spec
└── REMOVED Requirements → Delete the matching requirement from main spec

Merge carefully:

  • Match requirements by name (e.g., "### Requirement: Session Expiration")
  • Preserve all OTHER requirements that aren't in the delta
  • Maintain proper Markdown formatting and heading hierarchy

If Main Spec Does NOT Exist

The delta spec IS a full spec (not a delta). Copy it directly:

# Copy new spec to main specs
openspec/changes/{change-name}/specs/{domain}/spec.md
  → openspec/specs/{domain}/spec.md

Step 3: Move to Archive

IF mode is engram: Skip — there are no openspec/ directories to move. The archive report in Engram serves as the audit trail.

IF mode is none: Skip — no filesystem operations.

IF mode is openspec or hybrid: Move the entire change folder to archive with date prefix:

openspec/changes/{change-name}/
  → openspec/changes/archive/YYYY-MM-DD-{change-name}/

Use today's date in ISO format (e.g., 2026-02-16).

Step 4: Verify Archive

IF mode is openspec or hybrid: Confirm:

  • Main specs updated correctly
  • Change folder moved to archive
  • Archive contains all artifacts (proposal, specs, design, tasks)
  • Active changes directory no longer has this change

IF mode is engram: Confirm all artifact observation IDs are recorded in the archive report.

IF mode is none: Skip verification — no persisted artifacts.

Step 5: Persist Archive Report

This step is MANDATORY — do NOT skip it.

If mode is engram:

mem_save(
  title: "sdd/{change-name}/archive-report",
  topic_key: "sdd/{change-name}/archive-report",
  type: "architecture",
  project: "{project}",
  content: "{your archive report with all observation IDs for lineage}"
)

If mode is openspec or hybrid: the file was already written in Step 3.

If mode is hybrid: also call mem_save as above (write to BOTH backends).

Step 6: Return Summary

Return to the orchestrator:

## Change Archived

**Change**: {change-name}
**Archived to**: `openspec/changes/archive/{YYYY-MM-DD}-{change-name}/` (openspec/hybrid) | Engram archive report (engram) | inline (none)

### Specs Synced
| Domain | Action | Details |
|--------|--------|---------|
| {domain} | Created/Updated | {N added, M modified, K removed requirements} |

### Archive Contents
- proposal.md ✅
- specs/ ✅
- design.md ✅
- tasks.md ✅ ({N}/{N} tasks complete)

### Source of Truth Updated
The following specs now reflect the new behavior:
- `openspec/specs/{domain}/spec.md`

### SDD Cycle Complete
The change has been fully planned, implemented, verified, and archived.
Ready for the next change.

Rules

  • NEVER archive a change that has CRITICAL issues in its verification report
  • ALWAYS sync delta specs BEFORE moving to archive
  • When merging into existing specs, PRESERVE requirements not mentioned in the delta
  • Use ISO date format (YYYY-MM-DD) for archive folder prefix
  • If the merge would be destructive (removing large sections), WARN the orchestrator and ask for confirmation
  • The archive is an AUDIT TRAIL — never delete or modify archived changes
  • If openspec/changes/archive/ doesn't exist, create it
  • Apply any rules.archive from openspec/config.yaml
  • Return a structured envelope with: status, executive_summary, detailed_report (optional), artifacts, next_recommended, and risks
Weekly Installs
23
GitHub Stars
670
First Seen
Feb 17, 2026
Installed on
gemini-cli23
github-copilot23
codex23
amp23
kimi-cli23
opencode23