- name
SKILL.md
Beceri dosyası yayıncının deposunda kalır.
Kopyalanan komutu bir terminale yapıştırın. Bu SKILL.md dosyasını ilgili ajana kurar.
npx skills add mindscale-noah/MindMemOS --skill mindmemos-cli -a cursor -yKopyalanan komutu bir terminale yapıştırın. Bu SKILL.md dosyasını ilgili ajana kurar.
npx skills add mindscale-noah/MindMemOS --skill mindmemos-cli -a claude-code -yKopyalanan komutu bir terminale yapıştırın. Bu SKILL.md dosyasını ilgili ajana kurar.
npx skills add mindscale-noah/MindMemOS --skill mindmemos-cli -a codex -yKopyalanan komutu bir terminale yapıştırın. Bu SKILL.md dosyasını ilgili ajana kurar.
npx skills add mindscale-noah/MindMemOS --skill mindmemos-cli -a grok -yKopyalanan mesajı bir Grok Bot sohbetinde gönderin. Bot bu SKILL.md dosyasını GitHub'dan alır ve tüm Bot'larınızın paylaştığı becerilere ekler.
Install this skill: https://github.com/mindscale-noah/MindMemOS/tree/HEAD/skills/mindmemos-cliYayıncıdan
MindMemOS is a long-term memory layer for AI agents. The mindmemos CLI is the
integration surface: every memory operation is a subcommand that prints either a
human-readable line or, with --json, stable machine-readable output. Any agent
or script can drive memory by shelling out to it.
To connect memory to a specific agent host (e.g. an editor or assistant that
supports plugins), the host calls this same CLI. Host-specific install guides
live under references/ — see Host integrations.
The CLI ships as the Python package mindmemos-sdk and exposes a mindmemos
executable.
pip install mindmemos-sdk
# or, isolated so it's on PATH globally (recommended):
pipx install mindmemos-sdk
uv tool install mindmemos-sdk
Authenticate once. This writes a local config (API key, default user id, base URL). Operations that require a
user inherit the default user id, but memory search does not: omit --user-id for project-wide search or pass
it explicitly for user-scoped search.
mindmemos auth
# non-interactive:
mindmemos auth --api-key sk-... --user-id alice --base-url https://api.mindmemos.example.com
Verify:
mindmemos config show # masked key, base_url, user_id
mindmemos memory search "test" # confirms connectivity with a project-wide search
General shape: mindmemos <group> <command> [args] [options].
request_id and includes it in command responses for tracing.search / add support --json for stable machine-readable output (what scripts and host integrations parse).0 = success, 1 = API/config error, 2 = bad arguments. On non-zero exit the error text (including server stderr) is printed to stdout/stderr.Identity & scoping options (where accepted): --user-id (the human the memory
belongs to), --app-id, --agent-id, --session-id. Project isolation is
derived from the API key, not from these flags. For memory search, --user-id
is per-request and does not inherit the user configured by mindmemos auth.
mindmemos auth once.memory search to recall, memory add to store turns.memory get to inspect, memory update / memory delete to correct, memory feedback and memory dreaming to let the system consolidate.memory add — store new memoryExtracts durable facts from messages and persists them (with dedup/merge against existing memory).
# single line
mindmemos memory add --content "I'm allergic to peanuts" --user-id alice
# a conversation turn
mindmemos memory add --messages-json \
'[{"role":"user","content":"book me a window seat next time"},
{"role":"assistant","content":"Noted, window seats going forward."}]' \
--session-id sess-42 --json
# fire-and-forget
mindmemos memory add --content "prefers dark mode" --async
# with an image attached to the message
mindmemos memory add --content "look at this photo of my desk" --image ./desk.png --user-id alice
memory search — recall by relevanceUse before answering or acting when the agent needs prior user preferences, project facts, decisions, or past experience related to the current request.
mindmemos memory search "what are the user's dietary restrictions?" --top-k 5 --user-id alice
mindmemos memory search "travel prefs" --rerank --search-strategy agentic --token-budget 2000 --user-id alice --json
# project-wide search across all users in the API-key project
mindmemos memory search "project notes" --filter '{"mem_type":"fact"}'
# search by image, or include stored images in text search results
mindmemos memory search "cats" --query-image ./query.png --user-id alice
Multimodal mode: the image options above need the server project's
algo_config.multimodal.modeto bebasicorenhanced(per-project overridable). Inoffmode image sends are rejected withadd.images_unsupported/search.images_unsupported— tell the user to enable multimodal mode for the project instead of retrying.basicneeds the server'smm_embed_model_routerendpoints (when empty, images are stored but not searchable);enhancedalso uses the chat model as a VLM. Images are stored as S3 blobs — MinIO (started bymake dev/make db) is the only image backend;MINIO=0skips MinIO for the text-only layout — setalgo_config.multimodal.mode: "off"in that layout too (images always need the MinIO/S3 blob store).
memory get — list / filter (no query)Use for inspection, audits, dashboards, or manual curation when you need to enumerate stored memories rather than search by semantic relevance.
Returns memories in the current project, optionally filtered. Carries no actor identity — project scope comes from the API key.
mindmemos memory get --filter '{"app_id":"openclaw"}' --top-k 20
memory get-image — download a stored imageImage hits from memory search carry a short-lived image: URL (presigned S3/MinIO link, or a signed proxy URL — both download without an API key) plus metadata.blob_id / metadata.user_id. Use get-image to re-download by id when the link is absent or expired; --user-id must be the image owner and is verified by the server.
mindmemos memory get-image <blob_id> --user-id alice # writes ./{blob_id}{ext}
mindmemos memory get-image <blob_id> --user-id alice -o out.png
memory update / memory delete — correct by idUse memory update when a specific memory id is known and the stored content
should be rewritten because it is stale, incomplete, or partially wrong.
Use memory delete when a specific memory id is known and the memory should be
removed because it is invalid, duplicated, sensitive, or no longer appropriate.
mindmemos memory update mem_123 --content "allergic to peanuts and shellfish"
mindmemos memory delete mem_123 --yes
memory feedback — reinforce / correct memory qualityUse feedback after an outcome reveals whether recalled memory was helpful, missing, stale, or wrong; choose explicit or implicit mode based on whether the caller can provide the interaction context.
Feedback has two modes:
mindmemos memory feedback \
--text "the lunch recommendation was wrong; user dislikes spicy food" \
--messages-json '[{"role":"user","content":"I do not like spicy food."}]'
mindmemos memory feedback \
--text "the coffee preference was wrong" \
--messages-json-file turn.json \
--recalled-memories-json '[{"id":"mem_123","memory":"User prefers hot coffee."}]'
mindmemos memory feedback # omit --text: server analyzes recent adds
memory dreaming — consolidation passUse as a scheduled or background maintenance step to consolidate, merge, compress, or reorganize accumulated memories outside the hot request path.
mindmemos memory dreaming
mindmemos memory dreaming --sync --app-id openclaw
mindmemos auth / config show [--show-secret] / config reset [-y] — credentials & local settings.mindmemos skill <register|list|show|pull|evolve|push|update|rollback|history|diff|unregister> — SDK-managed skills. Use register <skill_dir_or_SKILL.md> --alias <alias> to save a local alias, then use that alias anywhere a skill id is accepted. Use push <skill> after editing local SKILL.md to upload a new version. Use update <skill|--all> [--yes] to checkout published heads, rollback <skill> --to <version_id> [--yes] to restore a cached/downloaded version after reviewing the replacement plan, and diff <skill> [--from <version_id>] --to <version_id> for a read-only unified diff.mindmemos memory add ... --skill-context-json '[...]' — optional explicit skill trace context. When omitted, the SDK has a best-effort fallback for OpenClaw-style SKILL.md tool-call text in the add messages; host integrations such as the OpenClaw plugin may still provide their own detection and pass this flag explicitly.mindmemos doctor — config/connectivity check.MindMemOS is a memory lifecycle, not just a key-value store. Pick the operation by intent:
Rules of thumb:
add + search are the hot path — almost every agent turn does one or both.feedback and dreaming are the slow path — they improve memory quality over time. feedback is event-driven (an outcome happened); dreaming is schedule-driven (periodic consolidation), not for a hot request path.update / delete / get are manual curation — fixing mistakes and inspecting state, usually by a human or an admin tool, not in normal conversation flow.--user-id (and --session-id where it matters) so memories don't leak across users.When memory operations live inside a Python agent/app rather than a shell call,
use the SDK shipped in the same mindmemos package (same API, same mindmemos auth config). See references/python-sdk.md for the
full sync + async example. Minimal sync usage:
from mindmemos_sdk import MindMemOSClient, DialogueMessage
with MindMemOSClient(user_id="alice") as client: # reads base URL and API key from `mindmemos auth`
client.memory.add(messages=[DialogueMessage(role="user", content="allergic to peanuts")])
hits = client.memory.search("dietary restrictions", top_k=5, user_id="alice")
for hit in hits.memories:
print(hit.id, hit.memory)
To wire MindMemOS into an agent host so memory is recalled and stored automatically (rather than calling the CLI by hand), follow the host-specific guide. All hosts depend on the CLI installed and authenticated above.
Give an AI agent persistent, cross-session long-term memory through MindMemOS. Covers installing and authenticating the mindmemos CLI, the full command interface (add / search / get / get-image / upda
---name: mindmemos-clidescription: Give an AI agent persistent, cross-session long-term memory through MindMemOS. Covers installing and…---# skills/mindmemos-cli/SKILL.mdGenel bakış
Give an AI agent persistent, cross-session long-term memory through MindMemOS. Covers installing and authenticating the mindmemos CLI, the full command interface (add / search / get / get-image / update / delete / feedback / dreaming, with image support on add and search) with parameters and examples, guidance on which capability to use when, plus a Python SDK example. To wire memory into a specific agent host (OpenClaw, DeepSeek Harness, Codex, Claude, etc.), see references/.
SSS
skills/mindmemos-cli/SKILL.md dosyasını yayıncıdan indirin veya yolu kopyalayıp depoyu açın. PlugMCP beceriyi listeler, dosyayı barındırmaz.
Yayıncı bunu şurada tutuyor: https://github.com/mindscale-noah/MindMemOS.
Claude, Cursor, Codex ve Grok Build, yayıncının deposundan eklediğiniz bir SKILL.md dosyasını okuyabilir. Grok Bot, becerinin GitHub klasör bağlantısını bir sohbette gönderdiğinizde onu kurar.
dreaming |
| An offline maintenance pass with no inputs. Run periodically (e.g. scheduled), not per-turn. |