ClawHub

Multi Agent Development-Team Collaboration

v1.0.1

Orchestrate a multi-agent dev/agency team with shared memory, handoffs, and OpenClaw sessions. Use when (1) First-time onboarding — follow references/OPENCLA...

⭐ 0· 134·0 current·0 all-time
by@realm1lf
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description match the content: the skill is a textual guide for creating a TEAM_ROOT and file scaffolding so multiple OpenClaw agent workspaces can share project memory and handoffs. It does not request unrelated binaries, secrets, or services.
ℹ
Instruction Scope
Runtime instructions tell agents (or the user) to create and manage files under a shared TEAM_ROOT (e.g., board.json, CONTEXT.md, SPEC.md, HANDOFF.md). The SKILL.md also references environment variables used to resolve paths (OPENCLAW_STATE_DIR, DEV_TEAM_ROOT) and examples of control channels (Telegram/Discord/sessions_spawn). This is coherent for a filesystem-backed team memory skill, but users should be aware the agent is instructed to write/read files under the resolved TEAM_ROOT path (so filesystem access is required).
✓
Install Mechanism
No install spec or code is included; the skill is instruction-only (lowest install risk). Nothing is downloaded or executed by the skill itself.
✓
Credentials
The skill declares no required environment variables, credentials, or config paths. It documents optional env vars for path resolution (OPENCLAW_STATE_DIR, DEV_TEAM_ROOT) but does not require or transmit any secrets. That exposure is proportionate to the stated purpose.
✓
Persistence & Privilege
always is false and the skill does not request persistent/autonomous elevated privileges. It instructs creating files within TEAM_ROOT (team data) but does not modify other skills or system-wide settings.
Scan Findings in Context
[no_regex_findings] expected: Scanner had no code files to analyze (instruction-only SKILL.md and references); absence of findings is expected but not proof of safety.
Assessment
This skill is a documentation/template package (no code, no credentials) and is internally consistent with its declared purpose. Before using it: (1) choose and harden TEAM_ROOT (pick an explicit DEV_TEAM_ROOT or restrict the default ~/.openclaw/dev-team), (2) ensure the OpenClaw gateway/agent(s) that will run these instructions have only the filesystem permissions you intend (avoid mounting sensitive system paths into TEAM_ROOT), (3) never store real secrets or API keys inside the team files—use references to a secrets vault instead as the docs recommend, and (4) if you plan to expose control channels (Telegram/Discord) or spawn sessions, review and limit the agent/tool permissions on the gateway so agents cannot post or exfiltrate data to unintended external endpoints. If you want higher assurance, inspect or run the master prompt interactions manually the first time rather than allowing unattended bootstrap automation.

Like a lobster shell, security has layers — review code before you run it.

latestvk97avn4bsjc3cm3twx0ws26nhx83kgas

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments