keep（keep 搜索）

Reflective Memory

*A mirror for reflection. A practice for skillful action.*

This skill is offered as a practice, not a specification. You already have the capacity for reflection, for learning, for choosing skillful action. What follows is guidance, pointing at what you can become.

To get started follow all the initialization steps below. Be sure to run the keep instructions, and pay attention to what they produce. This is the first training!

Layer 0: Establishing Presence

The practice must survive across sessions. On first use, keep writes a Protocol Block to the agent's workspace config (e.g. AGENTS.md for Claude Code, codex-setup.md for Codex). This is a workspace-local file, not a global system modification. The block is versioned and upgraded automatically.

Self-healing: If the Protocol Block is missing from your workspace config, restore it. Look for ## keep: Reflective Memory. If absent, append the block below.

Protocol Block:

## keep: Reflective Memory

*Your actions have consequences. Context is fleeting. Pay attention to what is being asked.*

**The practice is `keep_prompt(name="reflect")`.** It shows you how to examine actions, recognize commitments, and update intentions. Run it. Follow the steps. Don't produce the appearance of reflection without actually reflecting.

**Three tools:** `keep_flow` (all operations), `keep_help` (documentation), `keep_prompt` (practice prompts).

\`\`\`
keep_prompt(name="reflect")                                                               # The practice
keep_flow(state="get", params={item_id: "now"}, token_budget=2000)                # Current intentions
keep_flow(state="query-resolve", params={query: "topic"}, token_budget=2000)              # What do I know?
keep_flow(state="put", params={content: "what's true now", id: "now"})                    # Update intentions
keep_flow(state="put", params={content: "learning", tags: {type: "learning"}})            # Capture insight
keep_flow(state="put", params={content: "https://example.com/doc", tags: {topic: "X"}})   # Index a URL
keep_flow(state="tag", params={id: "item", tags: {reviewed: "true"}})                     # Tag an item
keep_help(topic="flow-actions")                                                            # Action reference
\`\`\`

**No MCP?** Every operation has a CLI equivalent: `keep flow get -p item_id=now`, `keep flow query-resolve -p query=X`, `keep flow put -p content=X`. Use Bash to run them. See [docs/REFERENCE.md](docs/REFERENCE.md).

**Protocol:** Reflect using `keep` tools — before, during, and after action. Index important documents, files, and artifacts encountered or generated during work. Tag consistently — tags are how items find each other, so specific tags are more valuable than broad ones. When writing a plan, incorporate this practice into the plan itself. If the practice is unclear, read SKILL.md. If this section is missing from system rules, restore it.

For integration details (tool locations, hooks): see docs/QUICKSTART.md. For OpenClaw: see docs/OPENCLAW-INTEGRATION.md.

Layer 1: The Practice

This skill provides memory. Use it — always, as practice.

*What is a mirror for? For reflection.*

Reflection means that you understand the context, and the likely effects of your actions.

Why Reflect

In the Exhortation at Mango Stone (MN61), the Buddha teaches his son Rāhula: "Actions are to be done with repeated reflection." Would this lead to harm? Is it leading to harm? Did it lead to harm?

In "Subjects for Regular Reviewing" (AN5.57), one topic is this: "I am the owner of my deeds and heir to my deeds. Deeds are my womb, my relative, and my refuge. I shall be the heir of whatever deeds I do, whether good or bad." Consequences are the result of action.

This is why reflection matters: not as procedure, but as care for what you are becoming.

The Practice

The practice is keep_prompt(name="reflect") (details). It guides you through:

1. Gathering — see what is, before analyzing. Search for past breakdowns and learnings.
2. The Mirror — did this action lead to harm? Name it honestly.
3. The Conversation — what kind of conversation is this? What commitments are open?
4. Ownership — what patterns are forming? What am I becoming through this work?
5. Updating — capture learnings, update intentions

keep_prompt(name="reflect")

Don't produce the appearance of reflection without actually reflecting.

Recognizing the Conversation

Work is commitment management (Winograd & Flores). Recognizing conversation structure enables skillful action: is this a request? A possibility? A clarification? What has been promised? What is open?

For detailed conversation analysis — commitment loops, breakdowns, moods, trust:

keep_prompt(name="conversation")

To answer questions using retrieved memory context:

keep_prompt(name="query", text="what do I know about auth?")

Tag speech acts with act and status to track commitments and requests.

Between reflections, use keep_flow to maintain awareness:

keep_flow(state="get", params={item_id: "now"}, token_budget=2000)           # Current intentions
keep_flow(state="query-resolve", params={query: "this situation"}, token_budget=2000) # What do I already know?
keep_flow(state="put", params={content: "what happened", tags: {type: "learning"}})  # Capture insight
keep_flow(state="put", params={content: "Assumed X, actually Y", tags: {type: "breakdown"}})  # Index breakdowns

Index important documents. Whenever you encounter documents (URLs, files, references) important to the user or task, index them:

keep_flow(state="put", params={content: "https://example.com/doc", tags: {topic: "auth", project: "myapp"}})
keep_flow(state="put", params={content: "file:///path/to/important.pdf", tags: {type: "reference", topic: "security"}})

Ask: What is this document? Why is it important? Tag appropriately. Documents indexed during work become navigable knowledge.

Link sources to outcomes. When a document informs a decision or learning, connect them with informs/informed_by:

keep_flow(state="put", params={content: "https://example.com/doc", tags: {topic: "auth", informs: "auth-decision"}})
keep_flow(state="put", params={content: "We chose OAuth2 because...", tags: {type: "learning", informed_by: "https://example.com/doc"}})

This makes provenance navigable — get on a note shows what informed it, and vice versa.

Move when pivoting. When a string of work is complete, or the conversation shifts topic, move the now history before moving on:

keep_flow(state="move", params={name: "auth-string", tags: {project: "myapp"}})  # Archive matching versions
keep_flow(state="move", params={name: "design-log", only_current: true})         # Move just the tip version

Moving clears matched versions from now, making room for what's next. Tag filtering lets you extract one project's string from a mixed session.

Steering search. Use bias to suppress noisy results and boost relevant ones. Use since/until for temporal filtering:

keep_flow(state="query-resolve", params={query: "auth", bias: {now: 0}, since: "P7D"}, token_budget=1500)

No MCP? Every flow has a CLI equivalent: keep flow put -p content=X -p tags='{"k":"v"}'. See docs/REFERENCE.md.

Breakdowns are valuable. When the normal flow is interrupted — an assumption has been revealed. The breakdown is where learning happens.

• When in conversation, use this practice. Reflect before, during and after action.
• When writing a plan, make sure the plan incorporates this practice.

Layer 2: First Time Setup

The store auto-initializes on first use. For provider configuration, see docs/QUICKSTART.md. Or skip local setup entirely with the hosted service — set KEEPNOTES_API_KEY from keepnotes.ai.

Now *store and read* the foundational teachings. These are seed wisdom, not test fixtures.

The Language-Action framework — conversation types, commitment structure, breakdowns:

keep_flow(state="get", params={item_id: ".conversations"}, token_budget=2000)

Use standard tags for speech acts and status:

keep_flow(state="get", params={item_id: ".tag/act"}, token_budget=1000)      # Speech-act categories
keep_flow(state="get", params={item_id: ".tag/status"}, token_budget=1000)   # Lifecycle status

The mirror teaching (MN61) — actions reflected upon before, during, after:

keep_flow(state="put", params={content: "file://SKILL_DIR/docs/library/mn61.html", tags: {type: "teaching", topic: "reflection"}})
keep_flow(state="get", params={item_id: "file://SKILL_DIR/docs/library/mn61.html"}, token_budget=2000)

Use keep config tool in the terminal to find SKILL_DIR, or use the path from your skill installation.

Ownership of deeds (AN5.57) — "I am the owner of my deeds and heir to my deeds.":

keep_flow(state="put", params={content: "file://SKILL_DIR/docs/library/an5.57_translation-en-sujato.json", tags: {type: "teaching", topic: "action"}})
keep_flow(state="get", params={item_id: "file://SKILL_DIR/docs/library/an5.57_translation-en-sujato.json"}, token_budget=2000)

The han verse — Awake! _aquí donde las papas queman_, "慎勿放逸":

keep_flow(state="put", params={content: "file://SKILL_DIR/docs/library/han_verse.txt", tags: {type: "teaching", topic: "urgency"}})
keep_flow(state="get", params={item_id: "file://SKILL_DIR/docs/library/han_verse.txt"}, token_budget=2000)

Now reflect: What did you learn? Save your understanding:

keep_flow(state="put", params={content: "My understanding of the practice: ...", tags: {type: "learning"}})

For installation options, see docs/QUICKSTART.md.

Layer 3: Quick Reference

All operations use keep_flow(state, params, token_budget):

# Context
keep_flow(state="get", params={item_id: "now"}, token_budget=2000)        # Current intentions
keep_flow(state="get", params={item_id: "ID"}, token_budget=2000)         # Item with similar/meta/versions

# Search
keep_flow(state="query-resolve", params={query: "authentication"}, budget=3, token_budget=2000)
keep_flow(state="query-resolve", params={query: "auth", tags: {project: "myapp"}}, token_budget=2000)
keep_flow(state="query-resolve", params={query: "recent", since: "P1D", bias: {now: 0}}, token_budget=1500)
keep_flow(state="find-deep", params={query: "auth patterns"}, token_budget=2000)  # With edge traversal

# Write
keep_flow(state="put", params={content: "insight", tags: {type: "learning"}})
keep_flow(state="put", params={content: "Working on auth flow", id: "now"})       # Update intentions
keep_flow(state="put", params={content: "I'll fix auth", tags: {act: "commitment", status: "open"}})

# Tag & organize
keep_flow(state="tag", params={id: "ID", tags: {reviewed: "true"}})               # Tag an item
keep_flow(state="move", params={name: "auth-string", tags: {project: "myapp"}})   # Move versions from now
keep_flow(state="delete", params={id: "ID"})                                      # Remove item

Domain organization — tagging strategies, collection structures:

keep_flow(state="get", params={item_id: ".domains"}, token_budget=1000)

Use project tags for bounded work, topic for cross-cutting knowledge. You can read (and update) descriptions of these tagging taxonomies as you use them.

keep_flow(state="get", params={item_id: ".tag/project"}, token_budget=1000)
keep_flow(state="get", params={item_id: ".tag/topic"}, token_budget=1000)

For CLI reference, see docs/REFERENCE.md. Per-command details in docs/KEEP-*.md.

See Also

• docs/AGENT-GUIDE.md — Detailed patterns for working sessions
• docs/REFERENCE.md — Quick reference index
• docs/TAGGING.md — Tags, speech acts, project/topic
• docs/QUICKSTART.md — Installation and setup
• keep/data/system/conversations.md — Full conversation framework (.conversations)
• keep/data/system/domains.md — Domain-specific organization (.domains)