Mpm Session Management logo

Mpm Session Management

Community
bobmatnyc
mpm-session-management

PM session management framework for context limits and session resume

Overview

Publisherbobmatnyc
Repositoryclaude-mpm
Skill namempm-session-management
Stars
152
Forks
34
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by bobmatnyc on GitHub. Read the source before you install it.

Installation

Install the Mpm Session Management AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/bobmatnyc/claude-mpm.git /tmp/claude-mpm
mkdir -p .claude/skills
cp -r /tmp/claude-mpm/plugin/skills/mpm-session-management .claude/skills/mpm-session-management
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Mpm Session Management in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Mpm Session Management on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Mpm Session Management is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

PM Session Management

Type: Framework Applies To: Project Manager Agent Load Priority: On-demand (context limit or session resume)

Umbrella skill. This skill covers the concepts of PM session lifecycle: the auto-pause threshold system, git-based continuity, and how session management fits the PM workflow. The concrete, user-invocable procedures live in two dedicated skills — use them for the actual steps:

  • mpm-session-pause — pause and persist current work state.
  • mpm-session-resume — load context from a paused session.

Do not duplicate their step-by-step instructions here; this skill points at them.

Purpose

This skill provides the conceptual model for PM session pause/resume when context limits are approaching or when resuming from a previous session. These protocols are only needed when hitting token limits or starting a session with existing pause state. For the operational steps, invoke mpm-session-pause / mpm-session-resume.

When This Skill Is Loaded

  • Context usage reaches 70%+ thresholds (informational warnings only)
  • Session starts with .claude-mpm/sessions/LATEST-SESSION.txt
  • User runs /mpm-session-resume

Context Usage Monitoring

The MPM framework continuously tracks context usage and provides informational warnings as usage approaches limits. Context-usage auto-pause (which used to abort sessions at 90% usage) was disabled in commit da5166603 (PR #646) to preserve legitimate active work.

Threshold Levels and Warnings

LevelUsageBehavior
Caution70%Informational warning: "Context usage at 70%. Consider wrapping up current work."
Warning85%Informational warning: "Context usage at 85%. Session nearing capacity."
Auto-Pause Disabled90%Informational warning: "Context usage at 90%. Consider wrapping up or running /compact (auto-pause is disabled)."
Critical95%Informational warning: "Context usage at 95%. Consider wrapping up or running /compact (auto-pause is disabled)."

Crossing these thresholds only produces informational warnings. No session pause is triggered, no ACTIVE-PAUSE.jsonl file is created. Sessions continue normally.

Manual Session Pause

To save session state before context limits:

  1. Use the /mpm-session-pause skill to explicitly create a session snapshot
  2. This captures git state, todos, task list, and accomplishments into .claude-mpm/sessions/session-{timestamp}.* files
  3. Sessions are stored in project-local .claude-mpm/sessions/ directory (not synced across machines)

Session Resume Protocol

Procedure. The step-by-step resume flow now lives in mpm-session-resume. Invoke that skill to load and display context from a paused session.

When resuming work:

  1. Use /mpm-session-resume to load the most recent paused session
  2. The skill detects .claude-mpm/sessions/LATEST-SESSION.txt and loads the associated snapshot
  3. Session accomplishments, pending tasks, and git context are displayed
  4. PM can continue work with full prior context

PM Response to Context Warnings

When PM sees context warnings (70%+), follow this protocol:

Immediate Actions

  1. Wrap up current work phase

    • Complete the current delegation cycle
    • Don't start new major tasks
  2. Document all in-progress tasks

    • Ensure all todos are updated with current status
    • Mark BLOCKED todos with specific blockers
    • Add context to in_progress todos
  3. Delegate remaining work with clear handoff

    • Provide detailed context to agents
    • Include acceptance criteria
    • Reference relevant files and commits
  4. Create summary

    • What was completed
    • What remains to be done
    • Any blockers or important context

Example Wrap-Up Sequence

Context at 85% - wrapping up current phase

Completed:
- ✅ OAuth2 implementation (commit abc123)
- ✅ Staging deployment verified

In Progress:
- 🔄 QA verification (api-qa testing login flow)

Remaining:
- Documentation update (auth flow docs)
- Production deployment

Creating session snapshot for clean resume...

Git-Based Session Continuity

Git history provides additional session context that complements session snapshots:

Useful Git Commands

bash
# Recent commits (what was delivered)
git log --oneline -10

# Uncommitted changes (work in progress)
git status

# Recent work (last 24 hours)
git log --since="24 hours ago" --pretty=format:"%h %s"

# Files changed recently
git log --name-status --since="24 hours ago"

Integration with Session Resume

When resuming a session, PM should:

  1. Load session snapshot (if available)
  2. Check git log for additional context
  3. Verify git status for uncommitted work
  4. Reconcile session state with git state

Example:

Resuming session...

Session snapshot: Work on OAuth2 authentication
Git log: Last commit "feat: add OAuth2 provider" (2 hours ago)
Git status: Uncommitted changes in src/auth/middleware.js

Context: Implementation complete, middleware updates in progress

Session Files Structure

Sessions are stored project-locally in .claude-mpm/sessions/:

.claude-mpm/sessions/
├── LATEST-SESSION.txt       # Pointer to most recent paused session
├── session-YYYYMMDD-HHMMSS.json   # Machine-readable session snapshots
└── session-YYYYMMDD-HHMMSS.md     # Human-readable markdown summary

File Purposes

  • LATEST-SESSION.txt: Points to most recent finalized session file
  • session-*.json: Complete session state (todos, git status, accomplishments, next steps)
  • session-*.md: Human-readable summary for quick review

Sessions are created manually via /mpm-session-pause skill. The ACTIVE-PAUSE.jsonl file (used during auto-pause) is no longer created since auto-pause is disabled.

Best Practices

When Context Limits Approach

  1. Monitor warnings: When you see context warnings (70%+), begin wrapping up
  2. Finish current phase: Complete the delegation in progress
  3. Update todos: Ensure all todos reflect current status
  4. Create session snapshot: Invoke /mpm-session-pause to save state for later resume
  5. Create handoff context: Document what's completed and what remains

When Resuming Sessions

  1. Review session snapshot: Understand what was accomplished
  2. Check git history: Verify actual state matches snapshot
  3. Validate uncommitted work: Any WIP that wasn't tracked?
  4. Continue from clear state: Don't duplicate completed work

Avoiding Session Bloat

  • Keep delegations focused and atomic
  • Don't load unnecessary context (use skills on-demand)
  • Complete and close todos regularly
  • Commit work incrementally (easier to resume)

Common Scenarios

Session pause/resume scenarios reduce to two procedures:

  • Approaching context limits / wrapping up → Follow the "PM Response to Context Warnings" protocol above, then invoke /mpm-session-pause to create a session snapshot that persists your work state.
  • Returning to work → Invoke /mpm-session-resume, which loads the most recent paused session snapshot and displays prior accomplishments, pending tasks, and git context.

Keep delegations atomic and commit incrementally so either procedure resumes from a clean point.

Integration with PM Workflow

Session management integrates with standard PM workflow:

User Request
[Check for session resume state]
Research (if needed)
[Monitor context usage — watch for warnings]
Implementation
[Approaching limit? → /mpm-session-pause]
Deployment
QA
Documentation
[Save session snapshot if context is high]
Report Results

Trigger Keywords

  • "context", "pause", "resume", "session"
  • "token", "limit", "usage"
  • "continue", "previous session"
  • Auto-loaded at 70%+ context usage
  • Auto-loaded when session files exist

Related Skills

  • mpm-session-pause - Procedure: pause and persist current work state
  • mpm-session-resume - Procedure: load context from a paused session
  • mpm-git-file-tracking - File tracking during pause/resume
  • mpm-verification-protocols - Verification state during pause
  • mpm-delegation-patterns - Resuming delegations mid-workflow

Frequently asked questions

What does the Mpm Session Management AI skill do?

PM session management framework for context limits and session resume

Why use Mpm Session Management on TypingMind?

Because you install it once and use it with any model. Mpm Session Management is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Mpm Session Management in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/bobmatnyc/claude-mpm/tree/main/plugin/skills/mpm-session-management. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Mpm Session Management?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Mpm Session Management?

As many as you like. As long as a model supports skills, you can use Mpm Session Management with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Mpm Session Management AI skill free?

It is published on GitHub by bobmatnyc. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇