Reference documentation for Causantic's Claude Code skills (slash commands).
Skills are installed by causantic init to ~/.claude/skills/causantic-<name>/SKILL.md. They work as slash commands in Claude Code and orchestrate the underlying MCP tools with structured prompts tailored to each use case.
Reconstruct how something happened — walks backward through causal chains ("how did we solve X?")
| Parameter | Required | Description |
|---|---|---|
query |
Yes | Natural language question about past work |
project |
No | Filter to a specific project slug. Defaults to the current project (derived from working directory) |
agent |
No | Filter to a specific agent (e.g., "researcher"). Seeds only; chains cross agents |
When to use: User asks about past work, previous decisions, errors solved before, or context from prior sessions. Before saying "I don't have context from previous sessions" -- always try recall first.
Example: /causantic-recall authentication implementation decisions
Broad discovery — find everything memory knows about a topic ("what do I know about X?")
| Parameter | Required | Description |
|---|---|---|
query |
Yes | What to search for in memory |
project |
No | Filter to a specific project slug. Defaults to the current project (derived from working directory) |
agent |
No | Filter to a specific agent (e.g., "researcher") |
When to use: Broad discovery, finding past context on a topic, as a starting point before using recall for deeper narrative.
Example: /causantic-search database migration patterns
Surface what came after similar past situations — walks forward through causal chains ("what's likely relevant next?")
| Parameter | Required | Description |
|---|---|---|
context |
Yes | Concise summary of the current task or topic |
project |
No | Filter to a specific project slug. Defaults to the current project (derived from working directory) |
agent |
No | Filter to a specific agent (e.g., "researcher"). Seeds only; chains cross agents |
When to use: At the start of complex tasks to check for relevant prior work, when encountering patterns that might have been solved before.
Example: /causantic-predict refactoring the auth module
Resume interrupted work -- start-of-session briefing.
| Parameter | Required | Description |
|---|---|---|
topic |
No | Topic to focus on, or time reference ("yesterday", "last week") |
When to use: Start of a session, user asks "where did I leave off?"
Example: /causantic-resume the API refactor
Replay a past session chronologically, or get recent history. Call with no arguments to get the most recent history up to the token budget.
| Parameter | Required | Description |
|---|---|---|
time range |
No | Natural language time reference ("yesterday", "past 3 days", "session abc123"). Omit for recent history. |
When to use: "What did I work on yesterday?", "Show me the last session", "What happened recently?", rebuilding context from a specific time period.
Examples:
/causantic-reconstruct— recent history (timeline mode)/causantic-reconstruct past 3 days/causantic-reconstruct yesterday
Discover available projects in memory with chunk counts and date ranges.
When to use: Before using project-filtered queries, checking what's been ingested, verifying memory coverage.
Example: /causantic-list-projects
Check system health and memory statistics.
When to use: After running causantic init to verify hooks are firing, when memory seems stale, diagnosing setup issues.
Output includes: Version, hook status, memory statistics, per-project breakdowns.
Example: /causantic-status
Multi-agent codebase review and cleanup plan.
When to use: Comprehensive codebase review combining code analysis with historical context from memory. Produces a plan (enters planning mode), not immediate changes. Spawns parallel specialist agents for thorough coverage without context exhaustion.
Specialists: Infrastructure (dependencies, security, linting, dead code), Design (architecture, code quality, duplication, testability), Documentation (inventory, accuracy, coverage, structure), Memory (decision history, tech debt, past attempts).
Phases: Reconnaissance (lead agent scans project), spawn specialists (4 agents in parallel), synthesis (cross-reference, deduplicate, resolve contradictions), write CLEANUP_PLAN.md.
Example: /causantic-cleanup
Gather deferred work, cleanup findings, and user goals into a phased roadmap.
| Parameter | Required | Description |
|---|---|---|
goal |
No | User-provided goals or feature descriptions |
When to use: After cleanup to organize next steps, when planning a release, when wanting to consolidate TODOs/tech debt/feature requests into a single prioritised document.
Sources: CLEANUP_PLAN.md backlog, existing ROADMAP.md, Causantic memory, codebase TODOs/FIXMEs, user-provided goals.
Phases: Gather candidates, deduplicate and classify, dependency analysis and ordering, present draft for human review.
Output: ROADMAP.md with items organized into Phase 0 (Foundation) through Phase 5 (Aspirational), plus a Deferred/Won't Do section.
Example: /causantic-roadmap add user authentication and improve test coverage
Delete memory by topic, time range, or session. Always previews before deleting.
| Parameter | Required | Description |
|---|---|---|
query |
No | Semantic query for topic-based deletion |
threshold |
No | Similarity threshold (0--1, default 0.6). Higher = more selective |
before |
No | Delete chunks before this ISO 8601 date |
after |
No | Delete chunks on or after this ISO 8601 date |
session_id |
No | Delete chunks from a specific session |
project |
Yes | Project slug (derived from cwd or asked) |
Workflow: Always previews first (dry-run), shows what would be deleted, waits for explicit user confirmation before deleting.
When to use: User asks to forget, remove, or clean up specific memory. Memory contains incorrect or outdated information.
Examples:
/causantic-forget authentication flow-- delete memory about authentication/causantic-forget everything before January-- time-based deletion/causantic-forget session abc12345-- delete a specific session
| User intent | Skill |
|---|---|
| "What do I know about X?" | search |
| "How did we solve X?" | recall |
| "Why does X work this way?" | recall |
| "What might be relevant?" | predict |
| "Where did I leave off?" | resume |
| "What did I work on yesterday?" | reconstruct |
| "What happened recently?" | reconstruct |
| "Review and clean up this codebase" | cleanup |
| "What should we work on next?" | roadmap |
| "Build a roadmap" | roadmap |
| "Forget/delete memory about X" | forget |
Skills are user-facing wrappers that guide Claude on how to use the underlying MCP tools. Users invoke skills via slash commands (e.g., /causantic-recall); the skill template instructs Claude on which MCP tool(s) to call, what parameters to pass, and how to format the output.
For direct MCP tool documentation, see the MCP Tools Reference.