Ejecuta cualquier Skill en Manus
con un clic

Ejecuta cualquier Skill en Manus con un clic

janitor-identify-environment

Estrellas1

Forks1

Actualizado21 de julio de 2026 a las 13:01

Report this Claude Code session's full runtime environment — terminal, OS, filesystem, sandboxing, CI, editor/IDE, network, cloud, python env, git/GitHub, MCP servers, wikimem, and installed plugins. Secret-safe and fail-open; writes full detail to disk as JSON and returns a compact digest. Use when the user asks where this session is running / what environment this is / which terminal, OS, container, or CI this is, or when you need to ground a decision in the actual runtime (network mount? headless? behind NAT? which tools installed?).

Instalación

Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.

Ejecutar en Manus

Fuente

Emasoft

Emasoft/ai-maestro-janitor

Abrir repositorio de GitHub Ver repositorios del creador

Descarga

Ejecutar en Manus

Ocupaciones relacionadasSOC

Basado en la clasificación ocupacional SOC

Desarrolladores de softwareOcupaciones informáticas y matemáticas·SOC 15-1252

SKILL.md

readonly

name

janitor-identify-environment

description

Janitor identify-environment

Overview

Pure observation of THIS session's runtime. It writes the full report to <repo>/reports/identify-environment/<timestamp>-env.json and prints only a compact digest + that path — so the caller's context holds the summary, not the whole object.

When to use

The user asks where/what this session is running (terminal, OS, container, CI, IDE).
You need a runtime fact before acting: is the project on a network mount? headless/background? behind NAT? which compilers/runtimes/tools are installed? which MCP servers are configured?

Instructions

Run the backing script and surface its COMPACT digest + the saved-path line (do NOT paste the whole JSON into the conversation):

uv run --script --quiet "${CLAUDE_PLUGIN_ROOT}/scripts/identify_environment.py" --online

Flags:

--online — enable the GitHub (gh) probes: branch-protection rulesets, repo metadata, the janitor's latest-release staleness check. OFF by default so a plain run makes no network call.
--json — print the raw machine-readable object to stdout (does NOT write a file).
--fast — skip the two slowest LOCAL probes (per-tool version strings + the listening-ports scan).

It reports, for THIS session: Terminal/program (by process-ancestry walk, not fragile $TERM_PROGRAM) + multiplexer + whether inside an ai-maestro agent; OS (system, version, kernel, arch, WSL); Filesystem (fstype, network-mount flag); Container/VM/sandbox (Docker/Podman/K8s/ WSL/Codespaces/devcontainer/Gitpod/flatpak/snap/firejail/app-sandbox + systemd-detect-virt); CI/remote (Actions/GitLab/CircleCI/Jenkins/Azure/Buildkite/Vercel/…); Editor/IDE + Claude Code (VS Code/Cursor/Windsurf/Zed/JetBrains; CLI vs desktop); Execution context (interactive TTY vs headless; linked worktree vs main checkout); Network (proxies credential-masked, VPN/ Tailscale/WireGuard/WARP, gateway, behind-NAT, DNS, firewall, interfaces, listening services); Cloud (AWS/Azure/GCP presence only, never a credential); Toolchain (compilers/runtimes/ package-managers/dev-tools with versions, incl. tldr-code/distill/fastedit/memgrep/lean-ctx, + python venv/conda/pyenv/uv); MCP servers (name + transport + scheme://host, token stripped); Git repo (remotes + GitHub slug, branches with dates/descriptions, active git hooks honoring core.hooksPath); GitHub --online (description/default-branch/visibility + branch-protection rulesets); Wikimem (LOCAL/PROJECT/USER note counts); Plugins (installed vs enabled, hook events, the janitor's own version + staleness, last upgrade timestamps); Claude auth (API-key vs OAuth); gh CLI (username + scopes); GitHub Actions (workflows + third-party actions used, Claude Code action present?, CI platforms); Releases --online (GitHub + PyPI/npm/crates.io presence + latest); Homebrew tap (flags the brew trust requirement); Fork/collaboration (fork + upstream); Repo topology (single vs mono-repo, single vs multi-git, single vs mixed language); User / PATH / launch chain.

Scope

Pure observation — it changes nothing (aside from writing its own JSON report), makes no network call except the opt-in --online GitHub probes, never emits a secret VALUE (key/token/secret/ password/credential reported as presence only; proxy + MCP endpoints credential-stripped; cloud creds presence-only; no -w keychain read), and fails open (an unavailable probe degrades to "unknown", never a crash).

Más de este repositorio

mismo repositorio

janitor-memory-consolidate

Emasoft/ai-maestro-janitor

CONSOLIDATE (MERGE) executor — fuses two duplicate memory notes about the SAME subject + same type into one page (dedup lessons, redirect backlinks, keep the oldest origin date), through the crash-safe transaction core, never editing a live page. ABSTAINS on uncertainty, cross-type/cross-scope/two-hub, or a third page sharing the subject. Use on a [janitor-memory-consolidate] marker, or "consolidate / merge / deduplicate the memory", "two notes cover the same thing", "fold these memory pages together".

2026-07-261

janitor-memory-recall

Emasoft/ai-maestro-janitor

RECALL — before working on a file or debugging/deciding, surface the right memory WIKI page and navigate it. Two entry points - FILE-anchored (about to edit a file → surface that functionality's HUB page and descend its links to the detail needed) and SYMPTOM ("have we hit this before?" → rank pages by how the symptom hits their description/title/tags). Use before re-deriving architecture/gotchas, or when the user says "recall memories about X" or "did we already solve this". Reads only the pages the task needs, degrading to grep when memgrep is absent. The RECALL leg of the wiki-memory protocol.

2026-07-261

janitor-memory-write

Emasoft/ai-maestro-janitor

MEMORIZE — capture a durable decision/fact into the project's memory WIKI as a navigable page, not a loose note. Use after a bug-autopsy gotcha, a design/architecture decision, a project constraint not derivable from code, a confirmed user preference, or when the user says "remember this", "memorize this", "save a memory". Finds the right existing wikimem page first so it never duplicates; only when none fits, creates a HUB / ASPECT / COMPONENT page via the expand/reduce decision, wires it into the See-also web, and indexes it by symptom. The MEMORIZE leg of the wiki-memory protocol.

2026-07-261

janitor-memory-update

Emasoft/ai-maestro-janitor

UPDATE — modify an existing memory WIKI page when something changes — ADD a decision to the page that owns the subject, CORRECT a memory a new discovery contradicts (the 2-step non-destructive protocol), or RESHAPE a page that outgrew its tier (expand a shared rule into an aspect, reduce element-specific detail into its component page). Use when a decision changes, a memory turns out wrong, a page mixes too many subjects, or the user says "update the memory", "that's no longer true", "fix the note about Y", "this memory is outdated". Keeps the See-also web, the hub map, and the lessons trail consistent. The UPDATE leg of the wiki-memory protocol.

2026-07-261

janitor-auto-manage-oauth-on

Emasoft/ai-maestro-janitor

Opts THIS machine into the janitor's unattended multi-account OAuth rotator. Sets an opt-in flag the always-on janitor daemon's 60s oauth-rotator-tick Task reads to rotate the live Claude Code credential to an alternate paid account before a rate-limit 429 stalls an overnight session. Default OFF, macOS only, idempotent; refuses if a pinning env var would defeat rotation. Trigger with /janitor-auto-manage-oauth-on, "turn on oauth rotation", or "enable account rotator".

2026-07-261

janitor-token-attribution

Emasoft/ai-maestro-janitor

Rank every project on this machine by how many tokens it is burning right now and name the one to throttle — the fleet token-attribution dashboard. The account 5h/7d utilization% is aggregate across all your parallel projects; this scans each project's transcripts to say WHICH one is over-consuming (large 5h share AND above its own baseline) and where its spike came from. Read-only. Use when the user asks which project is eating the token budget / which session to throttle / who is burning the rate-limit window.

2026-07-261

name

janitor-identify-environment

description

Janitor identify-environment

Overview

When to use

The user asks where/what this session is running (terminal, OS, container, CI, IDE).
You need a runtime fact before acting: is the project on a network mount? headless/background? behind NAT? which compilers/runtimes/tools are installed? which MCP servers are configured?

Instructions

Run the backing script and surface its COMPACT digest + the saved-path line (do NOT paste the whole JSON into the conversation):

uv run --script --quiet "${CLAUDE_PLUGIN_ROOT}/scripts/identify_environment.py" --online

Flags:

--online — enable the GitHub (gh) probes: branch-protection rulesets, repo metadata, the janitor's latest-release staleness check. OFF by default so a plain run makes no network call.
--json — print the raw machine-readable object to stdout (does NOT write a file).
--fast — skip the two slowest LOCAL probes (per-tool version strings + the listening-ports scan).