Every Soleri agent ships with 350+ operations across 20 engine modules out of the box. This page shows what your agent can do, grouped by facade.
For parameter details on common operations, see the API Reference. For CLI commands, see the CLI Reference. For term definitions, see the Glossary.
| Facade | Ops | What it does |
|---|
| Vault | 26 | Store, search, capture, and manage knowledge |
| Admin | 56 | Health checks, telemetry, plugins, packs, accounts |
| Chat | 41 | Session management, auth, voice, browser, notifications |
| Plan | 32 | Plans, grading, verification, evidence, reconciliation |
| Brain | 30 | Learning loop, pattern strength, recommendations, radar |
| Orchestrate | 26 | Lifecycle, projects, playbooks |
| Memory | 15 | Cross-session, cross-project, export/import |
| Agency | 15 | Proactive intelligence, file watching, pattern surfacing |
| Curator | 13 | Deduplication, health audits, enrichment, contradictions |
| Control | 13 | Persona, intent routing, modes, governance |
| Archive | 12 | Archival, lifecycle, temporal, knowledge maintenance |
| Operator | 10 | Operator profile learning, signals, adaptation |
| Loop | 9 | Iterative validation with convergence detection |
| Links | 9 | Zettelkasten linking, traversal, orphan detection |
| Sync | 8 | Git sync, Obsidian sync, pack synchronization |
| Intake | 7 | Content ingestion — books, URLs, text, batch import |
| Tier | 7 | Multi-vault tiers — connect sources, search across |
| Branching | 5 | Vault branching — create, list, merge, delete |
| Review | 5 | Knowledge review workflow — submit, approve, reject |
| Context | 3 | Entity extraction, knowledge retrieval, analysis |
| Domain Facades | 5 per domain | Domain-scoped CRUD + search |
Core knowledge storage, retrieval, capture, and sharing scope. Archival, sync, review, linking, branching, tier, and intake ops have moved to dedicated facades — see Archive, Sync, Review, Links, Branching, Tier, and Intake. Backward-compatible stubs remain on the vault facade for existing agents.
| Op | Auth | Description |
|---|
search | read | Search with two-pass retrieval. mode:"scan" for lightweight, mode:"full" for complete. |
load_entries | read | Two-pass retrieval pass 2 — load full entries by IDs from a previous scan. |
vault_stats | read | Entry counts by type, domain, severity. |
list_all | read | List entries with optional filters and pagination. |
vault_get | read | Fetch a specific entry by ID. |
vault_update | write | Update an existing entry. |
vault_remove | admin | Delete an entry. |
vault_bulk_add | write | Add multiple entries at once. |
vault_bulk_remove | admin | Remove multiple entries at once. |
vault_tags | read | List all tags in the vault. |
vault_domains | read | List all domains. |
vault_recent | read | Recently added or modified entries. |
export | read | Export vault entries as JSON intelligence bundles. |
| Op | Auth | Description |
|---|
capture_knowledge | write | Capture a pattern or anti-pattern with full metadata. |
capture_quick | write | Quick capture — title and description, auto-infer the rest. |
capture_enriched | write | Capture with LLM-enriched metadata (auto-tags, severity). |
search_intelligent | read | Semantic search with 6-dimension scoring. |
search_feedback | write | Rate a search result to improve future relevance. |
| Op | Auth | Description |
|---|
vault_import | write | Import entries from JSON. |
vault_seed | write | Seed entries from intelligence data files. |
vault_seed_canonical | write | Seed canonical entries from markdown files. |
vault_content_hash | read | Content hash for an entry (dedup key). |
vault_dedup_status | read | Deduplication status across vault. |
| Op | Auth | Description |
|---|
vault_detect_scope | read | Detect whether an entry is agent, project, or team scoped. |
vault_list_by_scope | read | List entries filtered by scope. |
vault_set_scope | write | Change an entry’s scope. |
System health, telemetry, plugins, packs, accounts, and diagnostics.
| Op | Auth | Description |
|---|
admin_health | read | System health check. |
admin_tool_list | read | List all registered facades and operations. |
admin_config | read | Current agent configuration. |
admin_vault_size | read | Vault storage size on disk. |
admin_uptime | read | Agent uptime since last start. |
admin_version | read | Engine and package versions. |
admin_reset_cache | admin | Clear all caches. |
admin_diagnostic | read | Full diagnostic report. |
admin_permissions | read | Current auth permissions. |
admin_module_status | read | Status of each loaded module. |
admin_env | read | Environment variables (sanitized). |
admin_gc | admin | Run garbage collection on stale data. |
admin_export_config | read | Export full agent configuration. |
admin_validate_instructions | read | Validate instruction files for quality. |
admin_persistence_info | read | Session store persistence status. |
admin_setup_check | read | Check setup completion status. |
admin_setup_run | admin | Run global setup (hooks, CLAUDE.md). |
admin_hot_reload | admin | Hot-reload agent configuration. |
admin_subsystem_health | read | Per-subsystem health status. |
admin_health_snapshot | read | Point-in-time health snapshot. |
| Op | Auth | Description |
|---|
admin_telemetry | read | Facade call telemetry. |
admin_telemetry_recent | read | Recent telemetry events. |
admin_telemetry_reset | admin | Reset telemetry counters. |
admin_vault_analytics | read | Vault usage analytics. |
admin_search_insights | read | Top missed queries, relevance scores. |
telemetry_errors | read | Error telemetry data. |
telemetry_slow_ops | read | Slow operation telemetry. |
| Op | Auth | Description |
|---|
llm_status | read | Available LLM providers and health. |
llm_rotate | write | Rotate to next API key in pool. |
llm_call | write | Make an LLM call with automatic retry and failover. |
render_prompt | read | Render a prompt template with variables. |
list_templates | read | List available prompt templates. |
admin_key_pool_status | read | Key pool status (size, active key, circuit state). |
| Op | Auth | Description |
|---|
plugin_list | read | List all registered plugins. |
plugin_status | read | Detailed status of a specific plugin. |
plugin_load | write | Load a plugin dynamically. |
plugin_activate | write | Activate a loaded plugin. |
plugin_deactivate | write | Deactivate a plugin. |
admin_list_plugins | read | List plugins with metadata. |
admin_plugin_status | read | Plugin health and configuration. |
| Op | Auth | Description |
|---|
pack_list | read | List available knowledge packs. |
pack_install | write | Install a knowledge pack. |
pack_uninstall | write | Uninstall a knowledge pack. |
pack_validate | read | Validate a knowledge pack before installing. |
| Op | Auth | Description |
|---|
admin_add_account | admin | Add an API account profile with encrypted storage. |
admin_remove_account | admin | Remove an API account profile. |
admin_rotate_account | admin | Rotate to a different API account. |
admin_list_accounts | admin | List all account profiles (keys never exposed). |
admin_account_status | read | Current active account status. |
admin_create_token | admin | Create a named API token with role-based access. |
admin_revoke_token | admin | Revoke an API token. |
admin_list_tokens | admin | List all tokens (names and roles only). |
| Op | Auth | Description |
|---|
admin_list_flags | read | List all feature flags. |
admin_get_flag | read | Get a specific flag value. |
admin_set_flag | admin | Set a feature flag. |
Session management, authentication, response chunking, MCP bridge, voice, notifications, browser isolation, and message queue for chat transports (Telegram, web, etc.).
| Op | Auth | Description |
|---|
chat_session_init | write | Initialize chat session management. |
chat_session_get | read | Get or create a chat session by ID. |
chat_session_append | write | Append a message to a session. |
chat_session_clear | write | Clear message history for a session. |
chat_session_delete | write | Delete a session entirely. |
chat_session_list | read | List all session IDs. |
| Op | Auth | Description |
|---|
chat_chunk_response | read | Split a long response into chunks for chat platforms. |
| Op | Auth | Description |
|---|
chat_auth_init | write | Initialize chat authentication. |
chat_auth_check | read | Check if a user is authenticated. |
chat_auth_authenticate | write | Authenticate a user with a passphrase. |
chat_auth_revoke | write | Revoke authentication for a user. |
chat_auth_status | read | Auth status — enabled, user count, list. |
| Op | Auth | Description |
|---|
chat_bridge_init | write | Initialize the MCP tool bridge. |
chat_bridge_register | write | Register a tool with the bridge. |
chat_bridge_list | read | List registered bridge tools. |
chat_bridge_execute | write | Execute a registered tool via the bridge. |
chat_compress_output | read | Compress verbose tool output for chat display. |
| Op | Auth | Description |
|---|
chat_cancel_create | write | Create an AbortSignal for a chat task. |
chat_cancel_stop | write | Cancel the running task for a chat. |
chat_cancel_status | read | Cancellation status — running tasks info. |
| Op | Auth | Description |
|---|
chat_update_init | write | Initialize the self-update manager. |
chat_update_request | write | Request a restart (self-update, rebuild, manual). |
chat_update_confirm | write | Clear restart context after successful startup. |
| Op | Auth | Description |
|---|
chat_file_detect_intent | read | Detect intent for a file — vision, text, or intake. |
chat_file_build_content | read | Build multimodal content from a file. |
chat_file_cleanup | write | Clean up temp files older than threshold. |
| Op | Auth | Description |
|---|
chat_notify_init | write | Initialize the notification engine. |
chat_notify_start | write | Start the notification polling loop. |
chat_notify_stop | write | Stop the notification polling loop. |
chat_notify_poll | write | Run all notification checks once. |
chat_notify_status | read | Notification engine status. |
| Op | Auth | Description |
|---|
chat_voice_transcribe | write | Transcribe audio using OpenAI Whisper. |
chat_voice_synthesize | write | Synthesize speech from text using OpenAI TTS. |
| Op | Auth | Description |
|---|
chat_queue_init | write | Initialize disk-based message queue. |
chat_queue_inbox | read | Read pending messages from inbox. |
chat_queue_reply | write | Send a reply to a queued message. |
chat_queue_drain | write | Drain outbox — read and remove pending responses. |
| Op | Auth | Description |
|---|
chat_browser_init | write | Initialize per-chat Playwright browser isolation. |
chat_browser_acquire | write | Get or create a browser session for a chat. |
chat_browser_release | write | Release a browser session. |
chat_browser_status | read | Browser session status — active sessions info. |
Multi-step task planning with grading, verification, evidence, and drift detection.
| Op | Auth | Description |
|---|
create_plan | write | Create a new plan with title, objective, and tasks. |
list_plans | read | List all active plans. |
get_plan | read | Get a plan by ID with full details. |
approve_plan | write | Approve a draft plan (2-gate system: plan then tasks). |
update_task | write | Update a task status within an executing plan. |
complete_plan | write | Mark an executing plan as completed. |
plan_iterate | write | Iterate on a draft plan to improve its grade. |
plan_split | write | Split a plan into executable tasks. |
plan_brainstorm | write | Design-before-code brainstorming with domain awareness. |
| Op | Auth | Description |
|---|
plan_reconcile | write | Compare what was planned vs what happened. Generates drift report. |
plan_auto_reconcile | write | Automated reconciliation with fast-path option. |
plan_complete_lifecycle | write | Extract knowledge from reconciled plan and archive. |
| Op | Auth | Description |
|---|
plan_grade | read | Grade a plan against quality criteria. |
plan_check_history | read | View grading check history for a plan. |
plan_latest_check | read | Get the most recent grading check. |
plan_meets_grade | read | Check if a plan meets a target grade. |
plan_auto_improve | write | Automatically improve a plan to meet a target grade. |
| Op | Auth | Description |
|---|
plan_submit_evidence | write | Submit evidence (test output, etc.) for task acceptance. |
plan_submit_deliverable | write | Submit a deliverable for a task. |
plan_verify_task | read | Verify task completion — checks evidence + reviews. |
plan_verify_deliverables | read | Verify all deliverables for a task. |
plan_verify_plan | read | Verify full plan completion — all tasks verified. |
plan_validate | write | Run post-execution validation checks. |
| Op | Auth | Description |
|---|
plan_review | read | Review a plan against quality criteria. |
plan_review_spec | write | Stage 1: spec compliance review. |
plan_review_quality | write | Stage 2: code quality review. |
plan_review_outcome | write | Record outcome of a review. |
| Op | Auth | Description |
|---|
plan_dispatch | read | Generate subagent dispatch instructions. |
plan_record_task_metrics | write | Record execution metrics for a task. |
plan_execution_metrics | read | Get execution metrics for a plan. |
plan_stats | read | Planning statistics: total, by status, rate. |
plan_archive | write | Archive a completed plan. |
High-level plan-execute-complete lifecycle, project management, and playbooks.
| Op | Auth | Description |
|---|
register | write | Register a project with the orchestrator. |
orchestrate_plan | write | Create an orchestrated plan with vault + brain context. |
orchestrate_execute | write | Start executing an orchestrated plan. |
orchestrate_complete | write | Complete with epilogue — evidence report, fix-trail quality signals, knowledge capture, session record. |
orchestrate_status | read | Current orchestration state. |
orchestrate_quick_capture | write | Quick-capture knowledge during orchestration. |
| Op | Auth | Description |
|---|
project_get | read | Get registered project details. |
project_list | read | List all registered projects. |
project_unregister | admin | Unregister a project. |
project_get_rules | read | Get project-specific rules. |
project_list_rules | read | List all rules across projects. |
project_add_rule | write | Add a project rule. |
project_remove_rule | admin | Remove a project rule. |
project_link | write | Link two projects (related, parent/child, fork). |
project_unlink | write | Remove a project link. |
project_get_links | read | Get links for a project. |
project_linked_projects | read | List all linked projects with details. |
project_touch | write | Update project last-accessed timestamp. |
Multi-step validated procedures stored in the vault.
| Op | Auth | Description |
|---|
playbook_list | read | List available playbooks, optionally filtered. |
playbook_get | read | Get a playbook with full steps and validation criteria. |
playbook_create | write | Create a new playbook with validated steps. |
playbook_start | write | Start executing a playbook. |
playbook_step | write | Execute the next step in a running playbook. |
playbook_complete | write | Complete a running playbook. |
playbook_match | read | Find playbooks that match a given context. |
playbook_seed | write | Seed playbooks from template data. |
The learning system that tracks pattern effectiveness, with a built-in learning radar and knowledge synthesizer.
| Op | Auth | Description |
|---|
brain_session_context | read | Current session context and active patterns. |
brain_strengths | read | Pattern strength scores — 4-signal scoring: usage, spread, success, recency. |
brain_global_patterns | read | Cross-domain pattern registry — patterns appearing across multiple domains. |
brain_recommend | read | Context-aware recommendations for current task with domain and source matching. |
brain_build_intelligence | write | Run full intelligence pipeline: strengths, global registry, domain profiles. |
brain_export | read | Export all brain intelligence data as JSON. |
brain_import | write | Import brain intelligence data from a previous export. |
brain_stats | read | Comprehensive brain statistics including intelligence pipeline stats. |
brain_decay_report | read | Temporal decay scores — reveals expiring, active, or expired entries. |
llm_status | read | LLM provider availability, key pool status, model routing config. |
| Op | Auth | Description |
|---|
brain_lifecycle | write | Start or end a brain session — tracks tools, files, plan context. |
brain_extract_knowledge | write | Extract knowledge proposals from a session using 6 heuristic rules. |
brain_archive_sessions | write | Archive completed sessions older than N days. |
brain_promote_proposals | write | Promote knowledge proposals to vault entries (governance-gated). |
brain_reset_extracted | write | Reset extraction status on sessions for re-extraction. |
session_list | read | List brain sessions with filters: domain, active, extracted. |
session_get | read | Get a specific brain session by ID. |
session_quality | read | 4-dimension quality score: completeness, artifacts, tools, outcome. |
session_replay | read | Replay a session — data, quality score, proposals, duration. |
| Op | Auth | Description |
|---|
record_feedback | write | Record feedback on a search result — accepted or dismissed. |
brain_feedback | write | Enhanced feedback with typed actions, source tracking, confidence, duration. |
brain_feedback_stats | read | Feedback statistics — counts by action/source, acceptance rate. |
rebuild_vocabulary | write | Force rebuild the TF-IDF vocabulary from all vault entries. |
Detects patterns from corrections, search misses, and workarounds. Auto-captures high-confidence signals, queues medium for review.
| Op | Auth | Description |
|---|
radar_analyze | write | Analyze a learning signal — routes by confidence to capture or queue. |
radar_candidates | read | Get pending radar candidates queued for review. |
radar_approve | write | Approve a pending candidate — captures it to vault. |
radar_dismiss | write | Dismiss one or more candidates — single ID or array. |
radar_flush | write | Auto-capture all pending candidates above a confidence threshold. |
radar_stats | read | Radar statistics — analyzed, captured, queued, dismissed, gaps. |
| Op | Auth | Description |
|---|
synthesize | read | Synthesize vault knowledge into briefs, outlines, talking points, or post drafts. |
Cross-session, cross-project knowledge persistence.
| Op | Auth | Description |
|---|
memory_search | read | Search across memory sources (patterns, sessions, identity). |
memory_capture | write | Capture a memory entry. |
memory_list | read | List memory entries. |
session_capture | write | Capture current session context. |
memory_by_project | read | List memories for a specific project. |
memory_deduplicate | write | Deduplicate memory entries. |
memory_delete | admin | Delete a memory entry. |
memory_export | read | Export all memories as JSON. |
memory_import | write | Import memories from JSON. |
memory_prune | write | Prune old or low-value memories. |
memory_stats | read | Memory usage statistics. |
memory_topics | read | Extract topic clusters from memories. |
memory_configure | write | Configure memory settings (extra paths, features). |
memory_cross_project_search | read | Search across all linked projects with weighted relevance. |
memory_promote_to_global | write | Promote a pattern to the global pool. |
Automated knowledge quality management.
| Op | Auth | Description |
|---|
curator_status | read | Curator health and configuration. |
curator_health_audit | read | Full vault health audit — duplicates, staleness, gaps. |
curator_detect_duplicates | write | Detect duplicate entries in the vault. |
curator_contradictions | read | Detect pattern vs anti-pattern conflicts. |
curator_resolve_contradiction | write | Resolve a detected contradiction. |
curator_hybrid_contradictions | read | Detect contradictions using hybrid (FTS + vector) search. |
curator_groom | write | Groom specific entries with updated tags. |
curator_groom_all | write | Groom all entries in a project vault. |
curator_consolidate | write | Run full consolidation (dedup, archive, contradictions). |
curator_enrich | write | Enrich entries with LLM-generated metadata. |
curator_entry_history | read | Change history for a specific entry. |
curator_record_snapshot | write | Record a point-in-time vault snapshot. |
curator_queue_stats | read | Pending enrichment and deduplication queue sizes. |
Agent persona, intent routing, operational modes, and governance.
| Op | Auth | Description |
|---|
get_identity | read | Current agent identity and guidelines. |
update_identity | write | Update agent name, role, or voice. |
add_guideline | write | Add a behavioral guideline. |
remove_guideline | admin | Remove a guideline. |
rollback_identity | admin | Rollback to a previous identity version. |
route_intent | read | Classify user intent (build, fix, review, plan, etc.). |
morph | write | Switch operational mode (build-mode, fix-mode, etc.). |
get_behavior_rules | read | Current behavior rules and constraints. |
| Op | Auth | Description |
|---|
governance_policy | read | View or set governance policies (quotas, retention). |
governance_proposals | read | List pending governance proposals. |
governance_stats | read | Governance metrics — approvals, rejections, quotas. |
governance_expire | write | Expire stale proposals. |
governance_dashboard | read | Full governance dashboard with all metrics. |
Iterative validation for convergence-driven tasks.
| Op | Auth | Description |
|---|
loop_start | write | Start a validation loop with mode and target. |
loop_iterate | write | Run one iteration and check convergence. |
loop_iterate_gate | write | Gate check — should iteration continue? |
loop_status | read | Current loop state, iteration count, progress. |
loop_cancel | write | Cancel an active loop. |
loop_history | read | Past loop runs with outcomes. |
loop_is_active | read | Check if a loop is currently running. |
loop_complete | write | Mark a loop as successfully completed. |
loop_anomaly_check | read | Check for anomalies in loop iteration data. |
Proactive file watching, pattern surfacing, warning detection, and proactive intelligence. When enabled, your agent monitors file changes and surfaces relevant vault patterns without being asked.
| Op | Auth | Description |
|---|
agency_enable | write | Enable agency mode — starts proactive file watching. |
agency_disable | write | Disable agency mode — stops watching and clears state. |
agency_status | read | Agency status — enabled, watching, detectors, pending. |
agency_config | write | Update watch paths, extensions, debounce, thresholds. |
agency_scan_file | read | Manually scan a file for warnings. |
agency_warnings | read | Get all pending warnings from recent file scans. |
agency_surface_patterns | read | Surface vault patterns relevant to a file change. |
agency_clarify | read | Generate clarification question when intent is ambiguous. |
| Op | Auth | Description |
|---|
agency_suggestions | read | Evaluate suggestion rules and return triggered proactive suggestions. |
agency_rich_clarify | read | Structured clarification questions with urgency, options, and implications. |
agency_suppress_warning | write | Suppress a warning by ID — prevents it from appearing in pending warnings. |
agency_unsuppress_warning | write | Remove suppression for a warning. |
agency_dismiss_pattern | write | Dismiss a surfaced pattern for 24h — prevents re-surfacing. |
agency_notifications | read | Drain pending notifications — returns and clears the notification queue. |
agency_full_status | read | Full status including suggestions, suppressions, dismissals, and notifications. |
Entity extraction, knowledge retrieval, and context analysis. Used internally by the orchestrator and available for direct use.
| Op | Auth | Description |
|---|
context_extract_entities | read | Extract named entities from a prompt — files, functions, domains, actions. |
context_retrieve_knowledge | read | Retrieve relevant knowledge from vault (FTS) and brain. |
context_analyze | read | Full context analysis — entities, knowledge, confidence, domains. |
Vault archival, lifecycle management, temporal metadata, and knowledge maintenance. These ops were previously part of the Vault facade and are still accessible there for backward compatibility. See Sync & Export for sync workflows.
| Op | Auth | Description |
|---|
vault_archive | write | Archive entries older than N days to the archive table. |
vault_restore | write | Restore an archived entry back to the active entries table. |
vault_optimize | write | Optimize vault database — VACUUM, ANALYZE, FTS index rebuild. |
vault_backup | read | Export the full vault as a JSON bundle for backup or transfer. |
| Op | Auth | Description |
|---|
vault_age_report | read | Entry age distribution — today, week, month, quarter. |
vault_set_temporal | write | Set valid_from and valid_until timestamps on an entry. |
vault_find_expiring | read | Find entries expiring within a given number of days. |
vault_find_expired | read | List expired entries (valid_until in the past). |
| Op | Auth | Description |
|---|
knowledge_audit | read | Audit vault quality — coverage gaps, staleness, tag health. |
knowledge_health | read | Health metrics — freshness, staleness, contradiction signals. |
knowledge_merge | write | Merge two similar entries — keeps best metadata from both. |
knowledge_reorganize | write | Re-categorize entries — rename domains/tags. Dry-run by default. |
Operator profile learning — tracks personality, preferences, expertise, and communication style across sessions. Builds an adaptive profile from accumulated signals. See Operator Learning for how the profile evolves.
| Op | Auth | Description |
|---|
profile_get | read | Get the full operator profile or a specific section. |
profile_update_section | write | Update a specific profile section with new data. |
profile_correct | write | Correct a profile section — takes a snapshot before overwriting. |
profile_delete | admin | Delete the operator profile (takes a snapshot first). |
profile_export | read | Export the operator profile as markdown or JSON. |
signal_accumulate | write | Accumulate operator signals for later synthesis. |
signal_list | read | List operator signals with optional type/processed filters. |
signal_stats | read | Signal statistics — counts by type, unprocessed total. |
synthesis_check | read | Check if a synthesis pass is due based on thresholds. |
profile_snapshot | write | Create a versioned snapshot of the current profile. |
Git synchronization, Obsidian bidirectional sync, and shareable pack import/export. These ops were previously part of the Vault facade. See Sync & Export for usage guides.
| Op | Auth | Description |
|---|
vault_git_push | write | Push vault entries to a git-tracked directory as JSON files. |
vault_git_pull | write | Pull entries from a git directory into vault with conflict resolution. |
vault_git_sync | write | Bidirectional sync — push vault entries, pull git-only entries. |
| Op | Auth | Description |
|---|
obsidian_export | read | Export vault entries to Obsidian-compatible markdown. |
obsidian_import | write | Import Obsidian markdown files with YAML frontmatter. |
obsidian_sync | write | Bidirectional sync with Obsidian (push, pull, or both). |
| Op | Auth | Description |
|---|
vault_export_pack | read | Export vault entries as a shareable intelligence pack with links. |
vault_import_pack | write | Import an intelligence pack with content-hash dedup and link import. |
Knowledge review workflow — team review with submit, approve, reject lifecycle. These ops were previously part of the Vault facade. See Knowledge Review Workflow for the full review process.
| Op | Auth | Description |
|---|
vault_submit_review | write | Submit a vault entry for team review. |
vault_approve | admin | Approve a pending entry (pending_review to approved). |
vault_reject | admin | Reject a pending entry with a comment. |
vault_pending_reviews | read | List all entries pending team review. |
vault_review_stats | read | Review workflow statistics — counts by status. |
Content ingestion — PDF books, URLs, raw text, and batch import. Extracts knowledge items via LLM classification and stores them in the vault with deduplication. See Content Ingestion for usage examples.
| Op | Auth | Description |
|---|
intake_ingest_book | write | Ingest a PDF book — parse, chunk, create a resumable job. |
intake_process | write | Process pending chunks — extract, classify via LLM, dedup. |
intake_status | read | Get job status or list all intake jobs. |
intake_preview | read | Preview extraction for a page range before full intake. |
| Op | Auth | Description |
|---|
ingest_url | write | Fetch a URL, extract text, classify via LLM, dedup, and store. |
ingest_text | write | Ingest raw text (article, transcript, notes) — classify and store. |
ingest_batch | write | Ingest multiple text items in one call with per-item metadata. |
Zettelkasten entry linking — create typed links, traverse the knowledge graph, find orphans, and auto-link entries using FTS or LLM evaluation. See Entry Linking & Knowledge Graph for the linking model.
| Op | Auth | Description |
|---|
link_entries | write | Create a typed link between two entries (supports, contradicts, extends, sequences). |
unlink_entries | write | Remove a link between two entries. |
get_links | read | Get all links for an entry (outgoing + incoming backlinks). |
traverse | read | Walk the link graph from an entry up to N hops deep (1-5). |
suggest_links | read | Find semantically similar entries as link candidates via FTS5. |
get_orphans | read | Find entries with zero links (orphan detection). |
relink_vault | write | Smart re-linking — LLM-evaluated batch linking with reasoning notes. |
backfill_links | write | Generate links for orphan entries using FTS5 suggestions. |
link_stats | read | Graph statistics — total links, by type, most connected, orphans. |
Vault branching — experiment with knowledge changes without affecting the main vault. Create branches, add operations, review, and merge. See Vault Branching for workflows.
| Op | Auth | Description |
|---|
vault_branch | write | Create a named vault branch for experimentation. |
vault_branch_add | write | Add an operation (add/modify/remove) to a branch. |
vault_branch_list | read | List all branches with entry counts and merge status. |
vault_merge_branch | admin | Merge a branch into the main vault. |
vault_delete_branch | admin | Delete a branch and all its operations. |
Multi-vault tiers — connect external knowledge bases, search across all sources with priority-weighted cascading.
| Op | Auth | Description |
|---|
vault_connect | admin | Connect an additional vault tier (project or team). |
vault_disconnect | admin | Disconnect a vault tier (cannot disconnect agent tier). |
vault_tiers | read | List all vault tiers with connection status and entry counts. |
vault_search_all | read | Search across all connected tiers with priority weighting. |
| Op | Auth | Description |
|---|
vault_connect_source | admin | Connect a named vault source with configurable priority. |
vault_disconnect_source | admin | Disconnect a named vault source. |
vault_list_sources | read | List all connected sources with their priorities. |
Each knowledge domain gets its own facade with 5 operations:
| Op | Auth | Description |
|---|
get_patterns | read | List domain entries filtered by tags and severity. |
search | read | Domain-scoped intelligent search. |
get_entry | read | Fetch a specific entry by ID. |
capture | write | Add a new pattern (with governance gating). |
remove | admin | Delete an entry from this domain. |
Domains are added with npx @soleri/cli add-domain <name>.