ppage
Captures session learnings, decisions, and context to a markdown file for future agent ramp-up. Use when user says "ppage", "page context", "save context", "capture learnings", or before ending a substantial work session.
Installation
Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.
View source ↗GuidesSkill instructions
Page Context
Synthesize the current session's important context, learnings, and state into a document that another LLM agent can use to immediately ramp up.
When to Use
- User explicitly invokes with "ppage" or similar
- Before ending a long or complex session
- When switching context to different work
- When user wants to preserve learnings
Process
- Reflect on the session: what was accomplished, what was learned, what decisions were made
- Identify key files that are central to the work
- Determine current state and logical next steps
- Note any open questions or blockers
- Capture constraints and preferences: Document any user preferences, project constraints, or environmental limitations discovered during the session
- Gather critical references: Note important external URLs, documentation, example data, or key values discovered
- Create Bootstrap Instructions: List the key files in priority order (most important first) with absolute paths so a future agent can read them to reach full context
- Generate a succinct but descriptive name for the session (2-4 words, kebab-case)
- Open TEMPLATE.md and follow it exactly
- Write output to
.ai/tmp/{YYYY-MM-DD}-{succinct-name}.md
Output Format
Use TEMPLATE.md as the output structure.
Guidelines
- Be concise: Target under 200 lines
- Be actionable: File paths should be exact, next steps should be concrete
- Be explicit: Don't assume the reading agent has any prior context
- Capture decisions: Include rationale so decisions aren't re-litigated
- Preserve constraints: User preferences and project constraints prevent re-litigation
- Name meaningfully: The filename should indicate what the session was about
- Bootstrap first: The ppage is a summary, not a full context dump. The Bootstrap Instructions tell the future agent which files to read to reach full context. Prioritize files (most important first), use absolute paths, and separate required vs optional files.
- Omit empty sections: If a section has no content (e.g., no constraints discovered, no external references), omit it entirely rather than leaving placeholders
Create the .ai/tmp/ directory if it doesn't exist.
Examples
Read the usage examples in the source repository.
View source ↗Source history
- Indexed on
- 2026-01-20
- Index updated
- 2026-02-04
- Branch
- main
Index and sync dates describe this directory. View the source repository for code history.
GitHub ↗chat Comments (0)
Sign in to join the discussion and leave a comment.
No comments yet. Be the first to share your thoughts!