harvest

name	harvest
description	Cross-rig knowledge consolidation. One-time sweep + ongoing tiered promotion. Walks all .agents/ directories, extracts learnings/patterns/research, deduplicates across rigs, and promotes high-value items to global hub. Triggers: "harvest", "consolidate knowledge", "cross-rig sweep", "knowledge federation", "harvest knowledge".
skill_api_version	1
user-invocable	true
context	{"window":"fork","intent":{"mode":"task"}}
metadata	{"tier":"knowledge","dependencies":[]}

Harvest — Cross-Rig Knowledge Consolidation

Sweep all .agents/ directories across the workspace, extract learnings, patterns, and research, deduplicate cross-rig, and promote high-value items to the global knowledge hub (~/.agents/learnings/).

What This Skill Does

The knowledge flywheel captures learnings per-rig, but they stay siloed. Harvest closes the loop by walking all rigs, extracting artifacts, deduplicating by content hash, and promoting high-confidence items to the global hub where every rig can access them via ao inject.

When to use: Before an evolve cycle, after a burst of development across multiple rigs, or weekly as part of knowledge governance.

Output: .agents/harvest/latest.json (catalog) + promoted files in ~/.agents/learnings/

Execution Steps

Implementation note: This skill drives the harvest workflow from inside Claude using the ao forge + ao extract primitives plus a manual promotion step. There is no ao harvest subcommand — it has been retired from the design (see ~/.agents/learnings/2026-05-21-ao-harvest-not-exists.md). Do not look for one.

Step 1: Extract from Recent Sessions

# For each recent session transcript:
ao forge transcript <session.jsonl> --queue --quiet
# Then bulk extract:
ao extract --all

Read the resulting queued items in ~/.agents/learnings/ and .agents/learnings/ and report:

• Sessions processed
• Total artifacts extracted
• Unique vs duplicate count (run ao dedup --merge to assess)
• Promotion candidates (artifacts with confidence >= 0.5 from frontmatter)

Step 2: Log and Proceed (Autonomous)

Default behavior: log the extraction summary ("Forge produced N artifacts from M sessions; M unique, K candidates for promotion.") and proceed directly to Step 3. Only if --review flag is set, surface the summary and ask:

Harvest will promote N artifacts from M rigs to ~/.agents/learnings/.
Proceed? [Approve / Adjust threshold / Abort]

Step 3: Promote Manually

Promotion is hand-driven (or shell-scripted) — there is no auto-promote subcommand. For each candidate:

# Inspect, then move to global hub:
cp .agents/learnings/<file>.md ~/.agents/learnings/

Ensure each promoted file has proper frontmatter (name, description, tier, confidence).

Step 4: Post-Harvest Cleanup

Run dedup on the promotion target to clean up any remaining duplicates:

ao dedup --merge ~/.agents/learnings/ 2>/dev/null || true

Step 5: Report Results

Report to user:

• Rigs scanned
• Artifacts extracted and unique count
• Duplicates found (with top duplicate groups)
• Artifacts promoted (with provenance)
• Top discoveries (highest-confidence cross-rig patterns)

Flags

Flag	Default	Description
--auto	off	Skip confirmation gate
--roots	~/gt/	Override root directories to scan
--min-confidence	0.5	Minimum confidence for promotion
--include	learnings,patterns,research	Artifact types to extract

Quick Start

/harvest                          # Full sweep with confirmation
/harvest --auto                   # Hands-free sweep
/harvest --min-confidence 0.7     # Only promote high-confidence items
/harvest --roots ~/gt/,~/projects/ # Scan additional directories

Governance

See references/governance.md for ongoing governance model: size budgets, sweep frequency, staleness thresholds, and cross-rig synthesis triggers.

See Also

• skills/athena/SKILL.md — Single-rig Mine/Grow/Defrag
• skills/flywheel/SKILL.md — Flywheel health monitoring
• skills/inject/SKILL.md — Knowledge injection into sessions
• skills/forge/SKILL.md — Transcript knowledge extraction

Reference Documents

• references/governance.md — Governance model for ongoing knowledge consolidation

Examples

Full sweep with confirmation

/harvest

Expected: Scans all configured roots, presents candidates for promotion, asks confirmation before writing to knowledge base.

Hands-free sweep

/harvest --auto

Expected: Promotes all items above default confidence threshold without prompting. Runs silently.

High-confidence only

/harvest --min-confidence 0.7

Expected: Only items with ≥70% confidence score are promoted; borderline items are skipped.

Scan additional directories

/harvest --roots ~/gt/,~/projects/

Expected: Scans both additional roots alongside default scan paths.

Troubleshooting

Problem	Cause	Fix
No items found	Scan roots empty or misconfigured	Verify --roots paths exist; run without --roots first to confirm defaults work
Duplicates promoted	Same item in multiple roots	Check governance.md for deduplication rules; run /athena defrag after harvest
Confidence scores all low	Shallow learnings with no cross-referencing	Write richer retro notes; run /forge first to extract more structured learnings
--auto promotes bad items	Confidence threshold too low	Raise to --min-confidence 0.8; review governance.md staleness thresholds
Knowledge base size exceeded	Too many promotions without cleanup	Run /flywheel health check; trim via governance size budget policy