| name | recall-session |
| description | Start or resume a work session. Use when: beginning a new conversation on an existing project, resuming after a break, onboarding to a codebase, or needing to establish what was done and what's next. Searches mem0 for project history, reads TODO.md, and produces a focused next-steps plan. |
| argument-hint | Project name or area of focus (e.g. 'auth service' or 'the whole project') |
Session Kickoff — Recall and Plan
Never start working blind. Spend 2 minutes recalling context before touching a single file. This prevents duplicate work, contradicting past decisions, and losing track of in-progress items.
When to Use
- First message in a new conversation about an ongoing project
- Resuming after any break longer than an hour
- Handing off work or starting a new phase
- "What was I doing?" — any time context is fuzzy
Procedure
Step 1 — Search mem0 for project history
mcp_wulfnet-mem0_mem0_get_tool_schema("search_memory")
Then:
tool_name: "search_memory"
tool_input: {
"query": "<project name or relevant keywords>",
"user_id": "tyler"
}
Look for:
- Past architectural decisions
- Known issues or blockers
- Conventions established (naming, stack choices, patterns)
- Last known state of the work
Step 2 — Read TODO.md
Read the workspace TODO.md to find:
- Unchecked items → what's still pending
- Recently checked items → what was just finished
- Blocked items → what needs attention
Step 3 — Check recent git activity (optional but recommended)
git log --oneline -10
git status
This tells you what files were last touched and if there's uncommitted work.
Step 4 — Think about the current state
mcp_wulfnet-think_think_invoke_tool
tool_name: "think"
tool_input: {
"thought": "Given what I know about this project from mem0 and TODO.md, what is the actual next priority? What blockers exist? What context does the user need before we start?"
}
Step 5 — Produce a session brief
Report back with:
- What was done (from mem0 + git log)
- Current blockers or open questions
- Recommended next action (single, specific, actionable)
- Any context the user should confirm (decisions that may have changed)
After the Session — Save to mem0
At the end of a productive session, persist what was learned:
tool_name: "add_memory"
tool_input: {
"content": "<what was built, decided, or discovered>",
"user_id": "tyler",
"metadata": { "project": "<project-name>", "date": "<YYYY-MM-DD>" }
}
Save:
- Architectural decisions and the reasons behind them
- Non-obvious gotchas discovered during implementation
- Which approach was tried and failed (so we don't try it again)
- Current status at end of session
Rules
- Do NOT start writing code before completing steps 1–4
- If mem0 returns nothing, that means this is a genuinely new project — proceed to
think-plan skill
- Session brief must be ≤ 10 bullet points — if you need more, you're not summarizing
- Always save to mem0 at session end — memory that isn't persisted is wasted work