agentleFS
KrishP147KrishP147/skills/sitrep/SKILL.md

sitrep

Skill37 starsChanged 47 days agoMITView on GitHub

Live status of this session in ≤10 lines — what's running, git state, PRs + CI, asked vs done, what needs the user, and a SAFE/NOT SAFE close verdict. Use for "how's it going", "status update", "what's running", "can I close this".

12 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 12 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read-only snapshot of right now.

    line 6
  2. Very specific:

    Running: every background Bash/PowerShell, Monitor, and background Agent launched this session without a completion notification yet.

    line 10
  3. Distinctive:

    Commit or push only if the yes named that too.

    line 40

“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “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

Read-only snapshot of right now; no file writes (progress-report covers history). The one exception is §4, through handoff-auto.

1. Gather

  • Running: every background Bash/PowerShell, Monitor, and background Agent launched this session without a completion notification yet. Per task: age + last line of its output file. ListAgents → note peers that look stalled.
  • Git, for each repo this session ran git in (not only the cwd's), per git worktree list entry: git status --porcelain plus git status --porcelain --ignored -- skilleddocs <handoff folder> (excluded handoffs; folder per AGENTS.md/CLAUDE.md, if set). Unpushed = git log <ref>.. --oneline, <ref> = @{u}, else origin/<branch> if it exists, else origin/<default> (<default> per the handoff-auto skill's TRACKING.md); never local <default> (on <default> it hides everything). No such ref → unpushed "unknown", counted as unpushed. 0 commits and clean → nothing to push; commits and no origin/<branch> → "never pushed". git symbolic-ref -q HEAD fails → detached: say "detached at <sha>"; git for-each-ref --contains HEAD refs/heads refs/remotes prints nothing → its commits are on no branch (lost on switch or worktree removal): NOT SAFE, and the fix is git switch -c <name>, never "push". Local branches not checked out: same unpushed rule (git for-each-ref refs/heads --format='%(refname:short) %(upstream:short) %(upstream:track)'); list each with commits.
  • PRs: gh pr list -R <o>/<r> --author @me --json number,title,url,isDraft,statusCheckRollup (<o>/<r> per the next skill's BOARD.md, Target repo section; on a fork also -R <parent>, where its PRs land) → passing / failing / pending.
  • Asked vs done: each distinct ask in this conversation → done / in progress / not started.
  • Needs you: keys, approvals, manual steps, unanswered questions. "none" if none.

2. Verdict

SAFE only when nothing runs and nothing is dirty or unpushed, or the mid-flight work is covered by a handoff: committed + pushed, or saved on disk per TRACKING.md: untracked, skilleddocs/ excluded (the only option on forks), or a dated handoff held uncommitted on a task branch in commit mode. That held file isn't dirty; never suggest committing it. Cover is for the work the handoff describes; other dirty/unpushed code still needs commit/push or a mention in the handoff. Commits on no branch are never covered.

An untracked handoff dies with its worktree: NOT SAFE if that worktree is not the main one (first git worktree list entry) or was created this session via pair/meta-orchestrator, or if unsure it survives. Reason names the handoff path: move it out or keep the worktree.

Else NOT SAFE + each reason.

3. Output (≤10 lines, no preamble, drop empty lines)

Running: 2 background (12m, "tests passing") | 1 agent pending
Git: 3 worktrees, 1 dirty, 1 unpushed (feat/x, 2 commits)
PRs: #12 passing | #14 CI failing
Asked → done: 4/5 (in progress: <what>)
Needs you: <item>
Verdict: NOT SAFE — stop feat/x's shell and push first

4. On "close" / "stop them"

  1. List exactly what will stop (task ids, agents) and anything that would be committed or pushed.
  2. Act on an explicit yes only; anything vaguer → repeat the list.
  3. TaskStop each (no tool → tell the user how).
  4. Commit or push only if the yes named that too.
  5. Mid-flight work → handoff-auto, print its path. Earns SAFE on forks and untracked repos too (§2 caveats apply).

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.