Back to Skills

github-elements-tracking

This skill should be used when the user asks to "track work across sessions", "create an epic", "manage issue waves", "post a checkpoint", "claim an issue", "recover from compaction", "coordinate multiple agents", "update memory bank", "store large documents", or mentions GitHub Issues as persistent memory, multi-session work, context survival, agent collaboration, SERENA MCP memory, or project-level context. Provides complete protocols for using GitHub Issues as permanent memory that survives c

4starsUpdated 3/7/2026

Security Assessment

Safe(100/100)
Security Score100/100

About github-elements-tracking

The github-elements-tracking skill provides a structured protocol for using GitHub Issues as a persistent memory system for AI-driven development workflows. It is designed to support long-running projects where context may be lost across sessions due to model limitations or context exhaustion. By treating GitHub Issues as durable storage, the skill enables work to continue across multiple sessions while preserving project-level knowledge, progress reports, and coordination between agents. The skill emphasizes strict adherence to user specifications through the "IRON LAW," which requires that all user instructions be followed exactly without modification, reinterpretation, or silent improvement.

The skill includes operational rules for agent reporting, project configuration, and controlled write boundaries. All agent activity reports must be recorded in two locations: the GitHub Issue thread and a project-local GHE_REPORTS directory, using a standardized timestamp-based naming convention. The system defines a fixed set of participating agents and enforces reporting consistency across them. It also integrates with a SERENA MCP memory bank to store project-level context and large documents, while respecting configuration options defined in a `.claude/ghe.local.md` settings file. These settings control enforcement levels, synchronization behavior, checkpoint reminders, Git worktree automation, and notification verbosity.

Typical use cases include multi-session development work, coordination between multiple agents, epic creation and issue wave management, posting checkpoints during long tasks, recovering project context after memory compaction, and maintaining persistent project documentation. The skill is intended for AI-assisted development environments that rely on GitHub Issues as a durable coordination layer and memory system. It is especially relevant for teams or automated agents that need consistent project state, structured reporting, and durable collaboration records that survive session resets or context limits.

FAQ

When should the github-elements-tracking skill be used?

It should be used when managing work that must persist across sessions, such as tracking epics, coordinating multiple agents, posting checkpoints, managing issue waves, recovering context after compaction, or using GitHub Issues as a persistent memory system.

Where must agent reports be stored?

All agent reports must be posted in two places: the GitHub Issue thread and the local GHE_REPORTS directory within the project. Both locations must contain the full report text.

How are reports named in the GHE_REPORTS directory?

Reports must follow the format `<TIMESTAMP>_<title or description>_(<AGENT>).md`, where the timestamp uses the `YYYYMMDDHHMMSSTimezone` format.

Does this skill support configuration or customization?

Yes. The skill reads project settings from `.claude/ghe.local.md`, which can control enforcement levels, enable or disable operations, configure checkpoint intervals, manage SERENA memory synchronization, and adjust notification verbosity.

Are there restrictions on where background agents can write files?

Yes. Background agents may only write to the project directory and its subdirectories, the parent directory for sub-git projects, `~/.claude` for plugin or settings fixes, and `/tmp`. Writing outside these locations is not allowed.

Install github-elements-tracking

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

  1. Copy the skill folder to .claude/skills/
  2. Claude will automatically detect and use the skill