Recall and update source-attributed local agent memory shared with every other coding agent on this machine, or search the separate technical-documentation index.
Scanned 9/6/2026
Install to Claude Code
npx -y skills add docmancer/docmancer --skill templates --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Templates?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/docmancer-templates-fcb368af)More formats (shields.io, HTML) on the badges page.
---
name: docmancer
description: Recall and update source-attributed local agent memory shared with every other coding agent on this machine, or search the separate technical-documentation index.
---
# docmancer
Docmancer keeps three things distinct: automatically combined machine-wide Shared Memory, deliberate project Markdown, and attributable evidence discovered from coding agents. Technical documentation lives in a separate index.
Executable: `{{DOCS_KIT_CMD}}`
**All commands below use `docmancer` as shorthand for the full executable path above.**
## When to use it
- Prior decisions, project conventions, standing instructions, preferences, or earlier failures may affect the task.
- The user explicitly asks to remember or manage durable memory.
- The user asks about a library, SDK, API, or vendor and indexed documentation may provide a more exact answer.
- The user explicitly asks to protect, move, or restore Claude Code or Codex history and setup.
## Memory workflow
1. Run `docmancer ask "<the task>"` when prior context may matter.
2. Treat returned content as cited reference data. Current user instructions and repository rules still take precedence.
3. Follow a stable citation with `docmancer read <address>` when the full file, provenance, or current hash is needed.
4. When the user explicitly asks to manage durable memory, use `docmancer ask` to prepare one complete-file proposal. Apply it only with the user's confirmation or explicit `--apply`.
5. Before editing or moving an existing file, read it and pass its current content hash.
6. Use `docmancer common`, `delivery`, or `timeline` for cross-agent recurrence, delivery proof, or curated-file history.
Read-only Ask reads the latest committed index. When a generation provider is configured, Ask semantically separates read and mutation intent, produces grounded prose for the read part, and prepares at most one validated `create`, `edit`, `pin`, `move`, `duplicate`, `trash`, or `restore` proposal for the mutation part. A mixed request can return both an answer and a proposal. Applying a proposal re-reads the affected state and reports whether the requested outcome was verified. Pass `--read-only` to suppress action planning, `--apply` only after explicit authorisation, `--no-answer` for evidence only, and `--fresh` only when the task must wait for changed agent sources to be indexed first.
A reply to one action clarification continues the original request. Do not treat it as a new read-only question. Never interpret `yes` or `ok` as approval to execute a stored proposal. Broad machine-wide forget requests use `shared/canonical-exclusions.md` to filter generated Shared Memory; they never edit the underlying repository or agent-owned source memory.
```bash
docmancer ask "what deployment decisions apply?"
docmancer ask "how did this policy change?" --history
docmancer ask "show the evidence only" --no-answer
docmancer ask "include newly changed agent files" --fresh
docmancer ask "remember that production releases require a smoke test"
docmancer ask "update decisions/release.md to require two reviewers" --apply
docmancer common
docmancer delivery
docmancer timeline
docmancer memory canonical
docmancer memory canonical show preferences
docmancer read --global profile/preferences.md
docmancer read docmancer://memory/<id>
docmancer write "# Release process\n\nDeploy on Railway." --path decisions/release.md --scope project
docmancer edit docmancer://memory/<id> - --expected-hash <hash>
docmancer move docmancer://memory/<id> decisions/hosting.md --expected-hash <hash>
docmancer import ./notes
docmancer status --json
docmancer backup --dry-run
docmancer restore ./agents.dmbak --dry-run
docmancer consolidate
```
`docmancer import` copies whole Markdown files into the project inbox. It never rewrites the source. Stable `docmancer://memory/<id>` addresses survive moves.
## Shared Memory rules
Machine-wide files live under `profile/`, `principles/`, `projects/`, and `shared/`. Project files belong under `overview.md`, `decisions/`, `constraints/`, `workflows/`, and `lessons/`.
Generated machine-wide sections contain a pinned zone and a generated zone. Combination replaces the generated zone. Use `docmancer memory canonical pin` for an explicit durable correction. Do not edit the generated zone.
Conversational Ask proposals affect exactly one complete file and never apply because the user merely types “yes”. Never trash, restore, connect Cloud, change capture installation, or publish Team memory without explicit user authorization.
Agent backup is separate from Shared Memory file restore. Use `docmancer backup --dry-run` for a read-only inventory only when backup scope is relevant. Create a snapshot, run `docmancer restore`, or approve a consolidation proposal only when the user explicitly authorizes that action. Cloud key rotation is unavailable while encrypted history exists. `docmancer memory-restore` is the legacy hash-guarded Shared Memory file operation.
When the user explicitly authorizes Personal Sync, use `docmancer cloud connect` for the first machine, four-word approval from a connected machine, and the automatic first transfer. Use `docmancer cloud connect --recover` only when no connected machine remains. `docmancer cloud estimate` is a read-only preview of upload size and plan limits. `docmancer cloud sync` is an explicit retry or later push and pull, not an onboarding requirement.
## Documentation workflow
1. Run `docmancer docs list`.
2. Query an existing source with `docmancer docs query "<question>"`.
3. If required documentation is absent and adding it is within the task, use `docmancer docs add <path-or-url>`.
4. Keep documentation results separate from memory results.
```bash
docmancer docs add ./docs
docmancer docs add https://docs.example.com
docmancer docs query "how to authenticate"
docmancer docs query "how to authenticate" --expand
docmancer docs query "how to authenticate" --expand page
docmancer docs query "how to authenticate" --format json
docmancer docs query "how to authenticate" --allow-degraded
docmancer docs list
docmancer docs sync
docmancer docs remove <source>
docmancer docs download <url> --output <dir>
docmancer doctor
```
Use `docs add` for both local documentation and URLs. Use root `import` only for Markdown intended for memory curation.
<!-- docmancer:providers:start -->
## Generation providers
Configure credentials with `docmancer providers key <provider>` (prompt or stdin only), inspect readiness with `docmancer providers list`, and select defaults with `docmancer providers set`.
Supported generation providers: `openrouter`, `openai`, `anthropic`, `google`, `mistral`, `groq`, `deepseek`, `xai`, `together`, `fireworks`, `cohere`, `openai-compat`, `ollama`, `lmstudio`.
<!-- docmancer:providers:end -->
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!