Sageox Summarize Team Activity
v0.3.0Generate an overall team summary covering the last 24 hours across all SageOx-enabled teams. Reads distilled daily entries via `ox distill history` and produ...
⭐ 0· 43·0 current·0 all-time
byGalex Yen@galexy
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the code and SKILL.md: the skill enumerates distilled history via `ox`, selects new entries, inlines them into a prompt, and calls `claude -p` to produce a Slack-ready summary. Required binaries (ox, claude, jq) and the provided install helpers align with that purpose.
Instruction Scope
Instructions stay within the stated task but explicitly instruct the agent to inline team entries into prompts for Anthropic's Claude, i.e., user data will be sent to an external model provider. The SKILL.md also requires validating any filesystem paths read from ~/.openclaw/memory before shell interpolation; the helper scripts assume they are invoked with validated/absolute paths. Confirm the agent/runtime enforces the validation steps described in SKILL.md.
Install Mechanism
Installs `claude` via the public npm package @anthropic-ai/claude-code and `jq` via Homebrew (expected). `ox` is installed by a bundled script that downloads a pinned release tarball from GitHub Releases and verifies a sha256 before extracting to $HOME/.local/bin. Download-from-GitHub-with-checksum is acceptable and coherent for this skill, but any binary-download step carries higher inherent risk than pure package installs — the script mitigates this with pinned tag/sha256 and a readiness gate.
Credentials
No per-skill env vars are requested. The skill will use whatever `claude` credentials are available (OAuth session or ANTHROPIC_API_KEY in the shell). This is proportionate but important: the distill entries (potentially sensitive) will be transmitted to Anthropic during summarization, so ensure you are comfortable with that and that ANTHROPIC_API_KEY (if used) is managed appropriately.
Persistence & Privilege
always:false and no system-wide privileged changes. The skill writes and reads its own state under ~/.openclaw/memory and installs ox to $HOME/.local/bin (user-level). It does not modify other skills' configs or request elevated privileges.
Assessment
This skill appears to do what it claims: gather recently-distilled SageOx entries and produce a Slack-ready summary via Anthropic Claude. Before installing, consider: (1) summaries will be sent to Anthropic — confirm you’re comfortable sending the team entries to that model provider and that any ANTHROPIC_API_KEY in your environment is deliberate; (2) the skill will download a pinned ox binary from GitHub and install it to $HOME/.local/bin and write state under ~/.openclaw/memory — review the install script if you want to verify the pinned release/checksums yourself; (3) SKILL.md requires the agent to validate filesystem paths read from memory files — ensure your OpenClaw runtime enforces that validation, since the helper scripts assume validated inputs; (4) the npm package @anthropic-ai/claude-code is installed for the CLI — verify the package source if you require extra assurance. If you accept those trade-offs, the skill is internally coherent.Like a lobster shell, security has layers — review code before you run it.
latestvk97c7bvs2ewd50h5p1z8mmp8yd84wyny
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
📰 Clawdis
OSmacOS · Linux
Binsox, claude, jq
Install
Install Claude Code CLI (npm)
Bins: claude
npm i -g @anthropic-ai/claude-codeInstall jq (brew)
Bins: jq
brew install jq