agentleFS
mustafahasanainmustafahasanain/frappe-workflow/skills/project-context/SKILL.md

project-context

Skill0 starsChanged 3 months agoMITView on GitHub

Create and maintain PROJECT_CONTEXT.md - the concise navigation and architecture map of a Frappe application. Full first-time analysis, incremental updates keyed to analyzed_commit, and strict update-decision rules.

33 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use code-review.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • GoogleSkill

    code-review

    Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.

    Also in 3 projects that use this

The file

Files
SKILL.md

Project Context Skill

Maintains docs/ai-context/PROJECT_CONTEXT.md in the target app repository: a concise map that prevents future agents from rereading the whole repository. It is not a code dump, not a Git history, not a feature archive, not a task tracker.

When to Use

  • During init when docs/ai-context/PROJECT_CONTEXT.md is missing → first-time analysis (references/full-project-analysis.md).
  • During init or start when it exists → incremental analysis (references/incremental-analysis.md).
  • During finalization when the approved task changed architecture or navigation → content update per (references/context-update-rules.md).

Inputs

  • Target app repository root (from bin/frappe-workflow detect --json).
  • Template: templates/project/PROJECT_CONTEXT.md (plugin repository).
  • Current HEAD (git rev-parse HEAD) and, for incremental runs, the recorded analyzed_commit.

Outputs

  • docs/ai-context/PROJECT_CONTEXT.md in the app repository, passing bin/frappe-workflow validate project-context.
  • Frontmatter with analyzed_commit set to the commit that was analyzed.

Preconditions

  • Detection succeeded; the app root is a Git repository.
  • For incremental runs: existing file parses (frontmatter present).

Stopping Conditions

  • Validation passes and analyzed_commit equals current HEAD → done.
  • The repository has no commits yet → record that, produce the structural sections from the working tree, set analyzed_commit after the first commit exists.
  • Existing file invalid/obsolete beyond repair → report it and perform one full re-analysis (this is the only case that rereads everything).

Prohibited

  • Copying implementation details that reading the referenced file provides.
  • Binding the app to a Site here (Site choice is task-level state).
  • Updating content when the update-decision question answers "no" (bumping analyzed_commit alone is allowed after confirming no content change is needed).
  • Inventing versions: frappe_version/erpnext_version come from the bench (apps.txt, __init__.py versions) or are recorded as Unknown.

Shared Rules

shared-workflow-rules.md, frappe-project-detection.md, file-lifecycle.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.