env-check
Skill8 starsChanged 36 days agoMITView on GitHub
Detect CLOUD_AGENT vs LOCAL_WORKSPACE and apply the matching agent constraints. Use at session start when choosing tests, credentials, Hub continuity fallback, or whether local skill CLIs (e.g. graphify) exist. Hub MCP unavailable (not runtime alone) triggers the forge `[ADHD]` issue mailbox; never invent Hub or graphify state.
- Reads credentials
3 other projects use a version of this.
Adopted by
Most recently taken up by
- allanninal/claude-code-skillstheir own version, first seen Oct 17, 2026
- NinjaSln-labs/agent-skillstheir own version, first seen Oct 2, 2026
- ericboy0224/learn-docker-and-k8stheir own version, first seen Mar 27, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- ericboy0224/learn-docker-and-k8s★ 472same name
- allanninal/claude-code-skills★ 1same name
- NinjaSln-labs/agent-skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository6 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/adhd-hub.mdcCursor rule
- .cursor/rules/graphify.mdcCursor rule
- graphifySkill
- adhd-hub-projectsSkill
- adhd-hub-sessionSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 3 are distinctive, and 5 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 49
Do not invent “I'm in cloud” from a single weak hint.
- Very specific: line 53
When RUNTIME_ENV: CLOUD_AGENT (Cursor Cloud / remote sandbox):
- Very specific: line 56
If a required local CLI is missing: one-line notice, then continue via repo tools (Read/Grep/Glob/tests) and any committed artifacts such …
- Distinctive: line 57
No native browser GUI or macOS/Windows desktop binaries.
- Very specific: line 58
Prefer injected environment / OIDC / platform secrets over committing or relying on .env.local
- Very specific: line 65
When RUNTIME_ENV: LOCAL_WORKSPACE (including local Docker / localhost):
- Very specific: line 69
Prefer Hub MCP (resolve_project, session_digest, upsert_progress, …) when it is reachable
- Distinctive: line 76
Prefer Hub MCP whenever it is available, on either runtime
“Common” is measured: kept by at least 6 in 10 of the 3 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.
Files
Env check — Cloud agent vs local workspace
When to use
Run early in a session (or when continuity / credentials / browser tests / local CLIs matter) to classify the host. Use the result as supporting context. The Hub continuity fallback is triggered when ADHD Hub MCP is missing, errored, unauthorized, or otherwise unavailable — not merely because the runtime looks like a cloud sandbox.
Quick detect
Resolve scripts/check_runtime.sh relative to this skill’s install location (not the ADHD Hub source tree alone).
Supported layouts:
| Layout | Example command |
| ADHD Hub source checkout | bash skills/env-check/scripts/check_runtime.sh |
Project-scoped install (.agents/skills) | bash .agents/skills/env-check/scripts/check_runtime.sh |
| Global / plugin install | From the installed env-check skill directory: bash scripts/check_runtime.sh |
# Project-scoped (common after adhd-hub sync-project / setup --project-skills):
bash .agents/skills/env-check/scripts/check_runtime.shOutput lines:
RUNTIME_ENV: CLOUD_AGENTorRUNTIME_ENV: LOCAL_WORKSPACERUNTIME_SIGNALS: …(container / Cursor markers; informational)HUB_CONTINUITY_TRIGGER: MCP unavailable …LOCAL_SKILL_CLIS: …(CLOUD_AGENT: do not assume; LOCAL_WORKSPACE: may be available)
Detection quality
The script does not treat $USER=root as cloud by itself (false positives on local root shells). It combines:
- Cursor Cloud markers:
CURSOR_AGENT,__CURSOR_SANDBOX_ENV_RESTORE,/tmp/cursor/cloud-agent-transcripts,/exec-daemon, plus agent sockets when paired with container cues - Container cues (
/.dockerenv, cgroup) as supporting signals only — local Docker alone staysLOCAL_WORKSPACE
Do not invent “I'm in cloud” from a single weak hint. Prefer the script output when unsure.
CLOUD_AGENT guidelines
When RUNTIME_ENV: CLOUD_AGENT (Cursor Cloud / remote sandbox):
- Do not assume machine-installed local skill binaries or CLIs exist (examples:
graphify, other opt-in companions installed on a developer laptop). Cloud/remote sandboxes often cannot see those tools even when project rules mention them. - If a required local CLI is missing: one-line notice, then continue via repo tools (Read/Grep/Glob/tests) and any committed artifacts such as
graphify-out/when present. Never fabricate graphify graph output or Hub continuity/progress/thread state. - No native browser GUI or macOS/Windows desktop binaries; prefer headless tests and CLI verification
- Prefer injected environment / OIDC / platform secrets over committing or relying on
.env.local - Local-only assumptions (opening
localhostGUIs, host Docker socket from the agent VM, Tailscale to a private Hub) may fail — plan accordingly - If Hub MCP tools are missing/errored/unauthorized: on the first substantial Hub-worthy turn, the first line of the reply MUST say Hub MCP is not available, plus a short fix hint (MCP URL → this Hub's
/mcp,ADHD_HUB_AUTH_TOKEN, restart the agent; skip/cancel Auth if it hangs until Hub OAuth is enabled). Never invent Hub state or claim a Hub write succeeded - Fallback still applies when Hub MCP is down: when issue-write access exists and the identity is on Hub Inbox authors, open/update a forge/GitHub issue titled
[ADHD] …with Goal / Focus / Next / Resume (or Now / Done / Next / Return). Optional labelsadhd-hub,project:<slug>,source:cursor— skip labels if the token cannot set them. Recommended: appendMade with [ADHD Progress Hub](https://github.com/uniskela/adhd-hub)under a non-imported heading (e.g.## Attribution). No secrets, private Hub URLs, internal hosts/IPs, absolute local paths, or transcripts. Otherwise state that continuity persistence is unavailable and continue authorized work.
LOCAL_WORKSPACE guidelines
When RUNTIME_ENV: LOCAL_WORKSPACE (including local Docker / localhost):
- Machine-installed local skill CLIs (e.g.
graphify) may be available — use them when present; still do not invent output if a binary is missing - Local Docker, localhost services, and host browsers are OK when the operator's machine provides them
- Prefer Hub MCP (
resolve_project,session_digest,upsert_progress, …) when it is reachable - Same no-hallucination rule if MCP is down: acknowledge + fix hint; never invent Hub state; use the forge
[ADHD]mailbox only when write access + Inbox authors allow it
Continuity priority
- MCP unavailable? → loud first-line warning + forge
[ADHD]fallback (if allowed) — this is the Hub continuity trigger (still true on CLOUD_AGENT) - Runtime env → choose tests, credentials, tooling constraints, and whether local skill CLIs are expected
- Prefer Hub MCP whenever it is available, on either runtime
See also: adhd-hub-session (full Hub protocol), docs/public/forge-issue-inbox.md.
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
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.

