agentleFS
cedaniel200cedaniel200/atlas-template/.agents/skills/session-startup/SKILL.md

session-startup

Skill3 starsChanged 20 days agoMITView on GitHub

Obligatory sequence at session start: read AGENTS.md, load phase skills, check memory, run tests and present a plan.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Present a plan to the user: before touching code.

    line 30
  2. Very specific:

    When switching roles, announce it as 🎩 Ahora soy <rol> before acting.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Session startup

The operational contract lives in AGENTS.md. This skill is the procedure.

Steps (in order)

  1. Read AGENTS.md — it is the entry point. If ## Commands, ## Structure or ## Do Not Touch are missing, stop and ask to complete them.
  2. Check the bootstrap state — if .atlas/state.json exists and status is pending, run .atlas/BOOTSTRAP.md before anything else.
  3. Resolve the OpenSpec command preference — read it from AGENTS.md (§ Preferences):

    • user or agent: no question.
    • ask: ask once at session start: "Do I run the OpenSpec commands this session, or do you?" and apply the answer to this session only.

    Combine it with the orchestration question if both are needed. archive and sync always require explicit approval.

  4. Load the phase skills from the Skill Registry in AGENTS.md. Do not load skills from packs that are not activated.
  5. Review the previous session — if persistent memory is configured, retrieve the context. If not, inspect the repo state (git status, git log -5).
  6. Run the tests with the command declared in AGENTS.md (## Commands). They must be green before writing code.
  7. Present a plan to the user before touching code.

Workspace check (two phases)

Phase 1 — at startup. Report the current branch, whether the working directory is a worktree and the active change. Informing creates nothing and asks nothing.

Phase 2 — before the first write (a change artifact, code, a QA report or documentation). Recheck the branch. If it is the base branch, isolate the change with the using-git-worktrees skill or stop and ask the human. A session that starts in read-only mode still repeats this check before it writes.

If something is missing

MissingAction
Test commandAsk; never invent one
"Do not touch" rulesAsk before editing
Bootstrap pendingRun the ritual first
Red testsReport and do not proceed
Writing on the base branchIsolate or stop; never write on it

Role announcement

When switching roles, announce it as 🎩 Ahora soy <rol> before acting.

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.