Skip to content

Garry's Opinionated OpenClaw/Hermes Agent Brain

TypeScript
26.9 KB
Raw Blame

MCP tool catalog

Every non-localOnly operation on the MCP surface: 179 tools across 27 areas. Starter marks membership in the ~41-op starter surface (src/mcp/surface.ts); Gate names the config key that must be true before remote callers see/call the op (gbrain config set <key> true). What a given token actually sees is further filtered per request by scope, bound-client fence, publish gates, and the per-client surface — see docs/operations/mcp-surface-runbook.md. Area names are non-contractual groupings.

admin

Tool Description Scope Starter Gate
get_brain_health Brain health as bands for agents without admin scope: write capacity (ok, low or critical, with a days band) and counts of write failures, ghost pages and unsearchable chunks as bands, for the sources you can read. read
get_health Brain health dashboard (embed coverage, stale pages, orphans) — remote callers see counters confined to their source grant. admin
get_stats Brain statistics (page count, chunk count, etc.) — remote callers see counters confined to their source grant. admin
get_status_snapshot Snapshot for gbrain status thin-client mode: sync freshness + last cycle + queue depths + worker liveness. admin
get_usage Aggregate chat usage + cost from the chat_usage_log ledger (per-model and per-phase token counts, cache reads/writes, USD estimates) with explicit coverage fields. admin
get_write_attribution Admin read: who created and who last changed a page, or one of its facts, takes or timeline entries. admin
mute_notice Mute a coaching or info notice (or first_run_decisions) for this client. write yes
quarantine_list List quarantined (hidden) and optionally content-flagged pages by scanning page frontmatter, newest-updated first. admin
run_doctor Run brain health checks and return a structured DoctorReport (thin-client doctor surface). admin
run_onboard Probe brain health + optionally submit onboard remediations. admin
run_skillopt Run SkillOpt against a single skill; its iterative loop (optimizer_mode "reflect") is deprecated. admin

advisor

Tool Description Scope Starter Gate
advisor Ranked, read-only "what to do next" for this brain: version drift, pending migrations, schema-pack issues, stalled jobs, usage-shape gaps, and setup smells. read mcp.publish_advisor

chronicle

Tool Description Scope Starter Gate
chronicle_day Life Chronicle: events + timeline entries on a given day (or its ISO week when week=true), ordered chronologically; each row backlinks to its depth page. read
chronicle_last_seen Life Chronicle: when an entity was last seen — its own timeline rows OR an event's who. read
chronicle_on_this_day Life Chronicle: events from the same calendar day in PRIOR years ("on this day"). read
chronicle_since Life Chronicle: events + timeline entries on or after a date, optionally filtered by event kind. read
volunteer_chronicle Life Chronicle agent-orientation: the recent timeline (last N days) + the current validity-resolved ontology for the named entities, in one zero-LLM payload, so an agent orients before acting. read

code

Tool Description Scope Starter Gate
code_blast BEFORE editing any function, run code_blast with the symbol name to surface every transitive caller grouped by depth (direct → 2-hop → 3-hop). read
code_callees When tracing how a function flows to its dependencies (DB calls, HTTP calls, file I/O), run code_callees from the entry point. read
code_callers BEFORE editing any function, run code_callers with the symbol name to find every caller (the people who'd be affected by your change). read
code_def Where is this symbol defined? read
code_flow When tracing how a request flows through the codebase from entry point to side effect (DB write, HTTP call, file I/O), run code_flow from the entry point. read
code_refs Find every reference to a symbol across the codebase (every file, every line). read

discovery

Tool Description Scope Starter Gate
request_tools More tools and params: no arguments lists your catalog; {tools: [names]} returns full schemas, every optional param included; {surface} widens it (per client or stdio session). read yes

entities

Tool Description Scope Starter Gate
entity_identity_list List cross-source entity identity groups and their member pages. read
extract_entities Extract entity names (people, companies) from text and create/update their brain stub pages. write
extraction_pending List what waits for review: unverified auto-extracted entity stubs (pending; promote or reject them with extraction_review), facts saved without an entity (unattributed; relink each with remember entity + replaces, never forget first), and, when include names it, recently created entity pages whose name is one edit or one acronym letter from another entity (possible_duplicates; flag only, never merged). read

facts

Tool Description Scope Starter Gate
purge_fact Owner-only, trusted local CLI on the brain host (gbrain forget <id> --purge). admin

identity

Tool Description Scope Starter Gate
get_brain_identity Brain identity + counters for thin-client banner — remote callers see counters confined to their source grant. read
whoami Your identity: transport, scopes and, over OAuth, client, source_id and federated_read. read yes

ingest

Tool Description Scope Starter Gate
get_ingest_log Recent ingestion log entries: what was already imported, and when. read yes
log_ingest Record one ingestion event (source type, reference, pages updated, summary) in the brain's ingest log. write

insights

Tool Description Scope Starter Gate
find_anomalies Anomalies in recent page activity, grouped by cohort (tag or type) against a baseline. read yes
find_contradictions Return suspected-contradiction findings from the most recent gbrain eval suspected-contradictions probe run, optionally filtered by slug and/or severity. read
find_experts Answers 'who in my brain knows about '. read
find_trajectory Return the chronological claim trajectory for an entity (typed metric values over time, plus auto-detected regressions and narrative drift). read
get_calibration_profile Read the active calibration profile for a holder. read
get_recent_salience Recently touched pages ranked by salience. read yes
volunteer_context Push-based context: volunteer brain pages relevant to a rolling conversation window WITHOUT being asked. read

jobs

Tool Description Scope Starter Gate
accept_wakeshift_terms Record that the user agreed to the agreement step a watch tool returned (agreement_required): this brain operator's terms and privacy policy and Wakeshift's Connector Terms and Privacy Policy. write mcp.publish_wakeshift
cancel_job Cancel a waiting, active or delayed job you may manage. admin yes
get_agent_job Poll a submit_agent job: status, result, queue_position. agent yes
get_job Get job status and details by ID. admin
get_job_progress Get structured progress for a running job. admin
get_job_stats Job queue statistics. admin
list_jobs List background jobs with optional status/queue/name filters; pass fields to project each job to the columns you need. admin
list_watches List what this connection is watching through Wakeshift (jobs, write requests and crons: open watches, plus ones that ended in the last day) with their state: open, delivered, dead or ended. write mcp.publish_wakeshift
pause_job Pause a waiting, active or delayed background job. admin
replay_job Replay a completed/failed/dead job, optionally with modified data admin
resume_job Resume a paused background job (back to waiting). admin
retry_job Re-queue a failed or dead background job. admin
send_job_message Send a sidechannel message to a running job's inbox (for jobs that read steering messages). admin
submit_agent Submit an agent job (agent scope; tools and budget bound to your client). agent yes
submit_job Submit a background job. admin
unwatch Stop watching: ends this connection's open Wakeshift watches on one job, write request or cron (or one watch by its id), so no wake comes for it. write mcp.publish_wakeshift
wait_job Wait until a job is completed, failed, dead or cancelled (at most 55 s per call), then return it like get_job plus terminal. admin
watch_cron This tool notifies you: it has gbrain wake this conversation through Wakeshift the next time the named cron reports its run (ok or failed) with gbrain wakeshift emit. admin mcp.publish_wakeshift
watch_job This tool notifies you: it has gbrain wake this conversation through Wakeshift when job id finishes (completed, failed, dead or cancelled), so you can stop polling. admin mcp.publish_wakeshift
watch_write_request This tool notifies you: it has gbrain wake this conversation through Wakeshift when your write request settles (committed, failed, conflict or cancelled), so you can stop polling get_write_request after write_pending. write mcp.publish_wakeshift

links

Tool Description Scope Starter Gate
add_link Create a typed link (edge) from one page to another in the same source. write
find_orphans Find disconnected pages. read
get_backlinks Links to a page, ≤500/call; group:"page" by referrer, newest first. read yes
get_links List a page's outgoing links (typed edges to other pages). read
list_link_sources Link provenances (e.g. read yes
rank_entities Rank pages by graph connections: the most-linked people, companies or other entities. read
remove_link Remove a link between two pages (optionally only one link_type or link_source). write
traverse_graph Walk the link graph from a page. read yes
wanted_pages Link targets that have no page yet, most-referenced first: each was written as a link but its page does not exist, so no edge exists. read

loops

Tool Description Scope Starter Gate
loops_close Close an open loop by id: status 'done' (handled) or 'dropped' (not going to). write
loops_mute Suppress a sender (email address) or thread id from opening NEW loops — the detector feedback primitive behind "never track this sender". write
loops_unmute Remove a sender/thread suppression added by loops_mute, so the detector can open NEW loops for it again. write
open_loops The open-loop engine's killer output: who is waiting on you, what you promised, and the context needed to respond. read

memory

Tool Description Scope Starter Gate
confirm_memory Owner only: mark a fact, take or page as confirmed by the user (trust tier user_confirmed). write
extract_facts Extract personal-knowledge facts (events, preferences, commitments, beliefs, ideas, and plain facts) from a conversation turn into the per-source hot memory. write
forget_fact Forget a fact by recording a durable withdrawal in its source and visibility. write
memory_propose Propose one timeline entry or link on a shared page for the person this call serves; the brain owner approves it before it is written. write
memory_recall Facts for the person this call serves, one per line: their own recent facts first, then facts matching query. read
memory_write Save facts for the person this call serves, readable by them alone. write

memory-settings

Tool Description Scope Starter Gate
get_memory_settings Memory writeback settings: the mode agents follow, the exact rule text, every mode's rule, background capture state, and whether you may change it. read
memory_conversation_mode Use when the user says "save everything from here" (all) or "stop saving" (off): returns this conversation's memory rule. write yes
set_memory_writeback Change the brain-wide writeback mode agents are told to follow (inherit uses the operator's setting). admin mcp.publish_memory_settings

memory-verbs

Tool Description Scope Starter Gate
context_pack Core memory, budget-packed cards, open threads and hot facts for up to 8 entities, zero LLM. read yes
delta Pages, facts, thread events changed since a cursor, zero LLM. read yes
entity Person/company/account card, zero LLM. read yes
forget Expire a remembered fact by its fact_id (never a page slug). write yes
recall Read facts by entity, since or session_id; query also searches pages. read yes
remember Save {fact, provenance (required), entity (subject; recall needs it)} or items[]. write yes
synthesize [EXPENSIVE / SLOW: LLM calls] Cited answer to a broad question across pages. read yes

ontology

Tool Description Scope Starter Gate
ontology_conflicts Life Chronicle: dimensions with ≥2 distinct current values from ≥2 provenances (genuine disagreement, not temporal supersession). read
ontology_dimensions Life Chronicle meta-ontology: which dimensions the brain tracks across entities, with entity + observation counts. read
ontology_get Life Chronicle: the current resolved per-entity ontology (dimension → value) at asof (default now), with provenance + confidence + validity. read
ontology_propose Life Chronicle: record one ontology observation (entity has dimension=value), sourced + confidence-weighted + bi-temporal. write

pages

Tool Description Scope Starter Gate
abort_page_upload Discard an uncommitted upload session: its staged parts are deleted and its quota is freed. write
append_page Add text to the end of a page, in time proportional to the text: use it for logs and conversations instead of re-putting a large page. write
begin_page_upload Start writing a page too large for one request: then upload_page_part from part 0 and commit_page_upload. write
cancel_write_request Cancel your accepted write before it publishes; returns the receipt. write yes
capture Quick note ("just remember this"): slugged under inbox/ by date + content hash, so recapturing is idempotent. write yes
commit_page_upload Write an upload's page once every part is sent; returns put_page's receipt. write
delete_page Soft-delete a page and remove its markdown file from the source working tree (the source local_path, or sync.repo_path when the source has none). write
edit_page Change part of a page: prefer this over put_page for small changes. write yes
fetch Fetch the full text of one search result by its opaque, source-qualified id (OpenAI deep-research contract: the search/fetch pair). read
find_pages Find pages by name or path (what ls/find would answer), sorted by slug. read
get_chunks Return a page's indexed content chunks (the units search ranks), in chunk_index order, 50 at a time by default. read
get_page Read a page; section/grep reads one part. read yes
get_page_upload An upload's state and next_part, to resume after a lost reply. write
get_raw_data Retrieve raw data for a page. read
get_versions Page version history, newest snapshot first: metadata only (id, snapshot_at, title, type, tags, frontmatter) for the newest 20 versions by default. read
get_write_request Your write's receipt by request_id (after write_pending or a lost reply). write yes
list_page_purges Owner-only (gbrain pages purges list). admin
list_pages List pages by type, slug_prefix, frontmatter where; what did I touch this week: sort=updated_desc. read yes
list_write_conflicts Pages whose writes keep ending in a source_changed conflict, grouped by slug. read
list_write_requests Your write receipts in one source, newest first (for a lost request_id). write yes
promote_guest_note Owner only: keep a guest note (a page in a guest source, origin "guest") as an owner page. write
put_page Complete content REPLACES the whole page: read get_page include_content:true; its revision is expected_revision. write yes
put_pages Write up to 50 complete Markdown pages (8 MB total) in one call; use instead of put_page for more than 3 pages. write
put_raw_data Store a raw provider payload (API response JSON) alongside a page, keyed by source. write
resolve_slugs Fuzzy-match a partial slug or title to page slugs. read yes
restore_page Restore a soft-deleted page (clear deleted_at) and re-create its markdown file on disk (the counterpart to delete_page removing it; the result write_through field reports the outcome). write
revert_version Restore a page to an earlier version from its history (a new revision; history is kept). write
unpurge_page Owner-only (gbrain pages unpurge <slug>). admin
upload_page_part Send the next part of an upload: part equals next_part, the whole call within part_max_frame_bytes, cut between code points. write
whois Exact identity lookup: entity pages whose email, title or declared alias equals an email address or name, ignoring case, accents and quotes. read

questions

Tool Description Scope Starter Gate
questions_list List pinned questions with freshness, blocked reason and next action, plus fresh/stale/pending counts. read
questions_pin Pin a question: gbrain keeps a cited answer current from your notes and flags sentences whose evidence changed. write
questions_refresh Refresh a pinned answer now (a paid model call): re-retrieves current evidence and edits the answer, or recomputes with full or after deletions. write
questions_status One pinned question: the answer with per-sentence stale flags, evidence watermark, last refresh, blocked reason, next action and verify step. read
questions_unpin Unpin a question: refreshes stop and its page is archived (your notes on it are kept). write

schema

Tool Description Scope Starter Gate
get_active_schema_pack Cheap identity packet for the active schema pack. read
list_schema_packs List installed schema packs (bundled + user-installed). read
reload_schema_pack Flush the in-process schema pack cache so the next loadActivePack re-reads from disk. admin
schema_apply_mutations Batched schema pack mutation. admin
schema_explain_type Resolved settings for a single page_type in the active pack. read
schema_graph Schema pack graph as JSON edges. read
schema_lint Lint the active (or named) schema pack. read
schema_review_orphans List pages with no active-pack type match. read
schema_stats Per-type page counts + typed-coverage from the DB. read

search

Tool Description Scope Starter Gate
assemble_evidence Deliver whole evidence for an ordered list of search hits (each {source_id, slug, chunk_id} from a prior search/query result): the same windows, sections or pages query returns with return_unit, packed into token_budget. read
cache_stats Semantic query-cache introspection: resolved knobs (enabled, similarity threshold, TTL) plus row counts and total hits. admin
query Hybrid search for concept/landscape questions and people/companies/projects by topic; adds fact rows. read yes
rate_answer Rate how useful an answer's retrieved evidence was, so this brain ranks better next time (zero LLM calls). write
search Cheap hybrid search, no LLM expansion, top 20: exact tokens, names, field values. read yes
search_by_image Image-as-query retrieval. read
search_exact Find every page containing a literal phrase or identifier, exhaustively: use instead of grep. read
search_modes Read-only search-mode dashboard: active mode, EVERY mode-bundle knob resolved with attribution (mode default vs config override), the three frozen bundles, and a reranker_readiness verdict (whether the resolved reranker will actually run; remote callers get the verdict without the host key inventory). read
search_stats Search observability over a window: cache hit rate, intent/mode mix, budget drops, rank-1 score drift, graph-signals failure counts. admin
search_tune Read-only tuning recommendations derived from the last 7 days of search telemetry: what should change, why, and the paste-ready config command per recommendation — relay them to the user. admin

skills

Tool Description Scope Starter Gate
delete_skill CAS-delete a canonical shared skill and revoke future managed activation. write + skill_editor yes
get_skill Fetch a skill's prose and follow it: when it says search or store, call the same-named MCP tool here. read yes mcp.publish_skills
get_skill_asset Read an approved file from an exact skill revision (data, never executed). read yes mcp.publish_skills
get_skill_policy Read the owner publication policy and CAS epoch, including when sharing is disabled. admin + skill_publisher
join_brain Enroll this authenticated principal to follow approved shared skills. read + skills_member_self yes
leave_brain Stop only this principal’s enrollment. read + skills_member_self yes
list_brain_skillpack Skillpacks this brain ships: skills and a scaffold spec to offer the user. read yes mcp.publish_skills
list_skills Skills: prose instruction sets (NOT executable code) for tasks with this server's tools, with triggers, usable_tools and unavailable_tools. read yes mcp.publish_skills
put_skill Publish a complete file-canonical skill revision with CAS and a durable receipt. write + skill_editor yes
set_skill_policy Explicitly approve a versioned shared-skill disclosure and follow policy. admin + skill_publisher
sync_brain_skills Get a complete authorized shared-skills view and optionally record an own issued-batch delivery acknowledgment. read + skills_member_self yes

sources

Tool Description Scope Starter Gate
sources_add Register a new source. sources_admin
sources_list List registered sources with page counts and remote_url. read
sources_remove Hard-remove a source (cascades pages/chunks/embeddings). sources_admin
sources_resolve_path Read-only: what gbrain did with each file path of a source, by its own rules. read
sources_status Per-source diagnostic. read

storage

Tool Description Scope Starter Gate
storage_find Find archived items on the storage shelf (files moved to B2, Supabase Storage, S3 and similar) by a plain phrase, such as "the October Gmail dataset". read
storage_get Read one card from the storage shelf by id: what the item is, every stored copy (store, bucket, path, restic repository and snapshot, size, hash, when verified), the original path, why it was archived, the restore form (restore.skill and restore.args, run by the agent's own restore skill; cards carry no commands or credentials), past restores, who registered it, and its revision. read
storage_list List cards on the storage shelf, sorted by id, optionally filtered by store, status or whether a copy is still on the source machine. read

tags

Tool Description Scope Starter Gate
add_tag Add one tag to a page (idempotent). write
get_tags List the tags on one page. read
remove_tag Remove one tag from a page. write

takes

Tool Description Scope Starter Gate
takes_add Record a take (typed claim) on a page: fact / take / bet / hunch, with a holder (who holds the belief: world, people/, companies/, or brain), weight 0..1, and optional source/since date. write
takes_calibration Calibration curve: resolved correct/incorrect bets binned by stated weight; observed vs predicted per bucket. read
takes_list List takes (typed/weighted/attributed claims) filtered by holder/kind/active/etc. read
takes_resolve Resolve a take: quality correct / incorrect / partial / unresolvable, with optional evidence text and measured value/unit. write
takes_scorecard Calibration scorecard for resolved bets: counts, accuracy, Brier (correct ∨ incorrect only), partial_rate. read
takes_search Keyword search across takes (pg_trgm similarity over claim text) read
takes_supersede Supersede a take with a replacement claim: the old row is struck through (kept for archaeology), the replacement appends at the next fence row number. write
takes_update Update a take's mutable fields (weight, source, since date). write
think Multi-hop synthesis across pages + takes + graph. read

timeline

Tool Description Scope Starter Gate
add_timeline_entry Append a dated entry to a page timeline. write yes
get_timeline Get timeline entries for a page, optionally filtered by date window. read