save

Session Save

Codex Compatibility

When running this skill in Codex, translate Claude Code-only primitives before acting: AskUserQuestion -> chat/request_user_input, TodoWrite -> update_plan, Task/TaskCreate/TeamCreate/SendMessage -> spawn_agent/send_input/wait_agent when available and allowed, and EnterPlanMode/ExitPlanMode -> a concise chat plan plus explicit approval. Resolve Read/Write/Edit/Bash/WebSearch/WebFetch to Codex file/shell/web tools, and map ~/.claude/... paths to ~/.agents/... or ~/.codex/... unless the task explicitly targets Claude Code.

Cursor Compatibility

When running this skill in Cursor Agent, translate Claude Code-only primitives before acting: AskUserQuestion -> AskQuestion; TodoWrite -> Cursor TodoWrite or an equivalent checklist; Task/TaskCreate/TeamCreate/SendMessage/multi-agent flows -> Cursor Task (subagents), parallel Tasks, or run_in_background when allowed (TeamCreate/SendMessage may have no exact match); EnterPlanMode/ExitPlanMode -> Plan mode (SwitchMode / CreatePlan) plus explicit user approval. Resolve Read/Write/Edit/StrReplace/Bash/web/search/MCP via Cursor Composer or Agent equivalents. MCP names written as mcp__server__tool typically map to call_mcp_tool with configured server identifiers. Map ~/.claude/... to ~/.cursor/skills/, .cursor/skills/, and .cursor/rules/ unless the task explicitly targets Claude Code.

Persist session context to Serena MCP memory for cross-session continuity.

<essential_principles>

Core Requirements

• All persistence uses Serena MCP tools exclusively (no agent-specific tools)
• Always check existing memories before writing to avoid overwriting valuable context
• Session checkpoint keys must include date: session_YYYY-MM-DD_<description>
• Pattern and learning memories use: pattern_<topic>
• Report what was saved as a structured summary to the user

</essential_principles>

Phase 1: Session Analysis

1. Review what was accomplished this session:
   • Files created or modified
   • Decisions made and their rationale
   • Problems encountered and solutions found
   • Tasks completed and tasks remaining
2. Identify what is worth persisting:
   • Session state: Current progress, next steps, blockers
   • Learnings: Reusable patterns, solutions to problems
   • Plans: Active plans or updated plans
   • TODOs: Outstanding work items

Phase 2: Memory Inventory

3. Call list_memories to see existing memories
4. Identify which existing memories need updates vs. new memories to create

Phase 3: Persist Session Checkpoint

5. Call write_memory with key session_YYYY-MM-DD_<summary>:

## Session: YYYY-MM-DD — <summary>

Agent: <Claude Code | Cursor | Codex> (<model>, if known)

### Accomplished
- [what was done]

### Decisions Made
- [product decision]: [rationale]
- (<Agent>) [tooling decision]: [rationale]

### Files Changed
- [file path]: [what changed]

### Pending / Next Steps
- [what remains to be done]

### Blockers (if any)
- [blocker description]

Agent: names the agent running this skill; add the model only when the agent reports it. Serena memory belongs to the project, not the agent, so Claude Code, Cursor and Codex all read the same checkpoints. A tooling decision (which browser / MCP / IDE tool to use, a model-specific workaround) binds only the agent that made it: prefix it with (<Agent>), e.g. (Cursor) browser QA runs through Cursor's built-in browser. A product decision (spec, design, API contract, naming) binds every agent: write it without a prefix.

Phase 4: Persist Learnings (if any)

6. For each reusable pattern discovered, call write_memory with key pattern_<topic>:

## Pattern: <name>

**Context**: [when this applies]
**Solution**: [the pattern/approach]
**Example**: [concrete example]
**When to Use**: [trigger conditions]

7. For each persistent TODO, call write_memory with key todo_<description>:

## TODO: <description>

**Priority**: [high/medium/low]
**Context**: [why this matters]
**Acceptance Criteria**: [how to know it's done]

Phase 5: Validation

8. Verify: session checkpoint created, learnings persisted, no critical context lost

Phase 6: Save Report

Report to the user:

## Session Saved

### Memories Written
| Key | Purpose |
|-----|---------|
| session_YYYY-MM-DD_xxx | Session checkpoint |
| pattern_xxx | [if any] |
| todo_xxx | [if any] |

### Next Session
Run `/load` to restore this context.

Memory Naming Conventions

See references/memory-conventions.md for the complete naming reference.

Quick summary:

Prefix	Purpose	Example
CRITICAL_*	Must-read rules	CRITICAL_activation_rule
session_YYYY-MM-DD_*	Session checkpoints	session_2026-02-09_auth-flow
plan_*	Active plans	plan_dark-mode
pattern_*	Reusable patterns	pattern_supabase-rls
discovery_*	Brainstorming results	discovery_api-options
todo_*	Persistent TODOs	todo_fix-login

Success Criteria

• Session accomplishments analyzed
• Existing memories checked (no accidental overwrites)
• Session checkpoint memory written with date-stamped key
• Agent: line written; every tooling decision carries its (<Agent>) prefix
• Learnings/patterns persisted (if any discovered)
• Save report presented to user