context-prime

name	context-prime
description	Loads structured project context at session start from README, docs, plans, git state, and memory MCPs. Use when orienting on a known project, resuming after a gap, or when the user says "prime context" or "/context-prime".

When to use

• orienting on a known project, resuming after a gap, or when the user says "prime context" or "/context-prime"

On-demand loading: Read this skill only when starting or resuming work on a known project. Do not load for unrelated tasks.

Context Prime Skill

Use this skill at the start of a session on a known project to quickly load structured context. Invoke when starting work, resuming after a gap, or when you need a fast orientation of the project state.

Trigger phrases

• "prime context"
• "load project context"
• "/context-prime"
• "orient me on this project"
• "what's the current state of the project"
• "context load"

Skill instructions

When this skill is invoked, execute the following steps to build and emit a structured context summary:

Step 1: Read project identity

Read these files if they exist (in order):

1. README.md — project overview, tech stack, quick start
2. package.json / pyproject.toml / Cargo.toml / go.mod — name, version, dependencies
3. .cursor/plans/workflow_state.md — current workflow phase and active tasks
4. AGENTS.md — agent conventions for this project

Extract:

• Project name and purpose
• Primary language(s) and frameworks
• Key scripts/commands (dev, test, build, lint)
• Current workflow state (if available)

Step 2: Read recent decisions and docs

Scan these paths for relevant context:

1. docs/ — architecture, decisions, API contracts (read top-level index or most recent files)
2. .cursor/plans/ — any .plan.md or .md files (read titles and summaries)
3. CHANGELOG.md or HISTORY.md — last 10 entries for recent changes

Step 3: Identify recently modified files

Run to find files modified in the last 7 days:

git log --since="7 days ago" --name-only --format="" | sort -u | head -20

Or use Glob to find recent files and read key ones for context.

Step 4: Check current git state

git status --short
git log --oneline -10
git branch --show-current

Extract:

• Current branch
• Recent commits (last 10)
• Uncommitted changes

Step 5: Query memory systems (if available)

If user-cursor10x-mcp is available, query for recent project context (e.g. getComprehensiveContext). If user-devcontext is available, query for active session state (e.g. initialize_conversation_context). Extract:

• Previous session context for this project
• Stored decisions and findings
• Active tasks or blockers

Step 6: Emit structured context summary

Output in this format:

## Project Context: <PROJECT_NAME>

**Purpose:** <one-line description>
**Stack:** <languages, frameworks, databases>
**Key commands:**
- Dev: `<command>`
- Test: `<command>`
- Build: `<command>`
- Lint: `<command>`

---

**Current branch:** <branch>
**Workflow phase:** <phase from workflow_state.md or "unknown">
**Active tasks:**
- <task 1>
- <task 2>

---

**Recent changes (last 7 days):**
- <file 1>: <what changed>
- <file 2>: <what changed>

**Uncommitted changes:**
- <file>: <M/A/D status>

---

**Architecture highlights:**
<2-3 sentences from docs/ or README>

**Open decisions / known issues:**
<from plans/ or recent commits>

---

**Suggested focus areas for this session:**
1. <area 1>
2. <area 2>

Notes

• This skill is inspired by the elizaOS /context-prime command pattern, adapted for Cursor skills format.
• If memory MCP tools are available, prefer querying them for richer cross-session context.
• Keep the output concise — aim for one screen of context, not an encyclopedia.
• For monorepos, scope to the relevant package/app directory when possible.