/compound-memory
Orchestrate governed persistent project memory across immutable evidence, llm-wiki, line-memory, reviewed session candidates, generated outputs, and bounded context packs.
Claude Code
$ npx -y skills add jmagly/aiwg --skill compound-memory --agent claude-code
Codex
$ npx skills add jmagly/aiwg --skill compound-memory -a codex -y
Cursor
$ npx skills add jmagly/aiwg --skill compound-memory -a cursor -y
Gemini CLI
$ npx skills add jmagly/aiwg --skill compound-memory -a gemini-cli -y
GitHub Copilot
$ npx skills add jmagly/aiwg --skill compound-memory -a github-copilot -y
OpenCode
$ npx skills add jmagly/aiwg --skill compound-memory -a opencode -y
Windsurf
$ npx skills add jmagly/aiwg --skill compound-memory -a windsurf -y
SKILL.md
compound-memory.SKILL.mdnamespace: aiwg
name: compound-memory
platforms: [all]
description: Orchestrate governed persistent project memory across immutable evidence, llm-wiki, line-memory, reviewed session candidates, generated outputs, and bounded context packs.
triggers:
- use compound project memory
- ingest sources into persistent memory
- review memory candidates
- maintain line memory and wiki
- build a memory context pack
- capture an output into memory
Compound Memory
Use this driver for the persistent compounding-memory pattern. It coordinates existing capabilities; it does not replace their storage, review, indexing, or promotion mechanics.
Start with readiness:
aiwg compound-memory status --json
Governed workflows
1. **Ingest** — preserve the source unchanged, calculate its identity and digest, then route long-form material through `memory-ingest` and llm-wiki. Route session transcripts through `aiwg sessions import` and `extract`. 2. **Retrieve/use** — query llm-wiki or the artifact index for linked detail; use `aiwg compound-memory context <task> --budget <characters> --json` to merge relevant line facts and wiki evidence. Add `--no-touch` for inspection. Never inject the complete stores into startup context. 3. **Write/capture output** — register the output and its source/context-pack lineage before proposing any extracted facts as session candidates. Run `aiwg compound-memory capture-output <file> ... --json` first, review the immutable digest and minimized references, then repeat with `--confirm --operation-id <preview-operation-id>`. This creates registration and index receipts only; it never promotes knowledge. 4. **Manage** — use exact-version candidate review and preview promotion to the declared `memory` or `line-memory` consumer. One reviewed candidate may produce independent receipts for both consumers. 5. **Review/update** — independently accept, reject, or defer proposed facts. Start from `aiwg compound-memory review --limit 50 --json`, then inspect and decide through the exact-version sessions commands. Workspace identity changes require a separate explicit review gate. For canonical context, run `aiwg compound-memory update <target> <key> <value> --source-ref <ref> --reviewer <id> --reason <reason> --json`, review its diff/conflicts, then repeat with the exact operation ID and `--confirm`. This never refreshes provider adapters. 6. **Maintain** — preview pruning, deduplication, contradiction repair, orphan repair, and index refresh. Confirm only the reviewed mutations. Preserve lifecycle tombstones and source-purge dependent dispositions. The compound command's confirmation may replay pending output registrations; all other actions remain delegated to their owning review workflows.
Until the compound-memory lifecycle ADR is accepted, do not invent a new automatic promotion path. Use the existing commands and receipts named above, and treat generated/model output as untrusted proposed knowledge rather than ground truth.
References
- @$AIWG_ROOT/agentic/code/addons/compound-memory/docs/overview.md
- @$AIWG_ROOT/agentic/code/addons/line-memory/skills/line-memory/SKILL.md
- @$AIWG_ROOT/agentic/code/addons/llm-wiki/skills/llm-wiki/SKILL.md
- @$AIWG_ROOT/agentic/code/addons/semantic-memory/skills/memory-ingest/SKILL.md
Read more
namespace: aiwg name: compound-memory platforms: [all] description: Orchestrate governed persistent project memory across immutable evidence, llm-wiki, line-memory, reviewed session candidates, generated outputs, and bounded context packs. triggers: - use compound project memory - ingest sources into persistent memory - review memory candidates - maintain line memory and wiki - build a memory context pack - capture an output into memory
Compound Memory
Use this driver for the persistent compounding-memory pattern. It coordinates existing capabilities; it does not replace their storage, review, indexing, or promotion mechanics.
Start with readiness:
aiwg compound-memory status --json
Governed workflows
1. **Ingest** — preserve the source unchanged, calculate its identity and digest, then route long-form material through `memory-ingest` and llm-wiki. Route session transcripts through `aiwg sessions import` and `extract`. 2. **Retrieve/use** — query llm-wiki or the artifact index for linked detail; use `aiwg compound-memory context <task> --budget <characters> --json` to merge relevant line facts and wiki evidence. Add `--no-touch` for inspection. Never inject the complete stores into startup context. 3. **Write/capture output** — register the output and its source/context-pack lineage before proposing any extracted facts as session candidates. Run `aiwg compound-memory capture-output <file> ... --json` first, review the immutable digest and minimized references, then repeat with `--confirm --operation-id <preview-operation-id>`. This creates registration and index receipts only; it never promotes knowledge. 4. **Manage** — use exact-version candidate review and preview promotion to the declared `memory` or `line-memory` consumer. One reviewed candidate may produce independent receipts for both consumers. 5. **Review/update** — independently accept, reject, or defer proposed facts. Start from `aiwg compound-memory review --limit 50 --json`, then inspect and decide through the exact-version sessions commands. Workspace identity changes require a separate explicit review gate. For canonical context, run `aiwg compound-memory update <target> <key> <value> --source-ref <ref> --reviewer <id> --reason <reason> --json`, review its diff/conflicts, then repeat with the exact operation ID and `--confirm`. This never refreshes provider adapters. 6. **Maintain** — preview pruning, deduplication, contradiction repair, orphan repair, and index refresh. Confirm only the reviewed mutations. Preserve lifecycle tombstones and source-purge dependent dispositions. The compound command's confirmation may replay pending output registrations; all other actions remain delegated to their owning review workflows.
Until the compound-memory lifecycle ADR is accepted, do not invent a new automatic promotion path. Use the existing commands and receipts named above, and treat generated/model output as untrusted proposed knowledge rather than ground truth.
References
- @$AIWG_ROOT/agentic/code/addons/compound-memory/docs/overview.md
- @$AIWG_ROOT/agentic/code/addons/line-memory/skills/line-memory/SKILL.md
- @$AIWG_ROOT/agentic/code/addons/llm-wiki/skills/llm-wiki/SKILL.md
- @$AIWG_ROOT/agentic/code/addons/semantic-memory/skills/memory-ingest/SKILL.md
Reusable project context and specialist workflows for the AI tools you already use. Plan software, coordinate specialist reviews, prepare campaigns, investigate incidents, organize research, curate media, and maintain operational knowledge.
Other skills on aiwg.
More skills from other plugins.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…
ask-matt
Ask which skill or flow fits your situation. A router over the skills in this repo.
Questions about compound-memory.
- Which coding agents can use compound-memory?
- compound-memory installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.

