← Back to Skills Marketplace
ogp-agent-comms
by
latentgenius
· GitHub ↗
· v0.6.0
· MIT-0
213
Downloads
1
Stars
1
Active Installs
6
Versions
Install in OpenClaw
/install ogp-agent-comms
Description
Interactive wizard to configure agent-to-agent communication policies (updated for multi-framework `--for` workflows, OGP 0.2.24+ peer identity, and 0.2.28+...
Usage Guidance
This skill appears coherent and limited to configuring OGP agent-comms policies. Before using it: (1) verify and audit the referenced npm package (@dp-pcs/ogp) and its source (GitHub repo) before running npm install -g; (2) run the ogp commands manually first (ogp config show, ogp federation list) to inspect current state and avoid unintended changes; (3) back up your OGP configuration before making policy changes; (4) pay attention to session-key and human-delivery settings (hooks.allowRequestSessionKey and humanDeliveryTarget) — only enable request of session keys if you trust the peer and your platform configuration; and (5) if you are unsure about the package, perform installs in an isolated/dev environment or consult the package's repository and publisher identity first.
Capability Analysis
Type: OpenClaw Skill
Name: ogp-agent-comms
Version: 0.6.0
The skill is an interactive configuration wizard for the Open Gateway Protocol (OGP) agent-to-agent communication system. It provides structured instructions for an AI agent to guide users through managing communication policies, peer identities, and message routing using the `ogp` CLI tool. While it involves executing shell commands and configuring network-related delivery targets (e.g., Telegram), these actions are well-documented and strictly aligned with the stated purpose of managing federated agent communications without any evidence of malicious intent or unauthorized data exfiltration.
Capability Assessment
Purpose & Capability
The name/description match the instructions: the SKILL.md exclusively documents CLI interactions with the OGP daemon (ogp commands) to configure agent-comms policies. No unrelated credentials, binaries, or system paths are requested.
Instruction Scope
Instructions limit activity to querying and updating OGP configuration, listing peers, and running the OGP 'agent-comms' interview. The document does not instruct the agent to read unrelated files, sweep environment variables, or exfiltrate data to third-party endpoints. It references platform hooks (e.g. /hooks/agent) only as part of normal behavior.
Install Mechanism
The skill itself has no install spec, but the README suggests installing a third-party npm package (@dp-pcs/ogp) to get the ogp CLI. npm is a standard distribution method (moderate risk); users should verify the package source and trustworthiness before running global installs.
Credentials
The skill declares no required environment variables or credentials and its runtime instructions do not request unrelated secrets. It references configuration flags (e.g. hooks.allowRequestSessionKey) that are OGP platform settings, which are appropriate to this domain.
Persistence & Privilege
The skill is instruction-only, has always:false, and does not request permanent platform presence or modify other skills' configurations. It instructs the user to run OGP commands that update OGP config (expected behavior for a configuration wizard).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install ogp-agent-comms - After installation, invoke the skill by name or use
/ogp-agent-comms - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.6.0
Refresh agent-comms guidance for current multi-framework and delegated-authority behavior.
v0.2.1
Updated for OGP 0.2.24: peer IDs now shown as public key prefixes (v0.2.1 bump).
v0.2.4
v0.2.4: scope negotiation, policy system, updated for OGP v0.2.3
v0.2.3
- Documentation update: corrected Stanislav's peer hostname in the multi-select example for peer configuration.
- No functional or behavioral changes to the skill.
v0.2.2
No user-visible changes.
- No file changes were detected in this version.
- Documentation and functionality remain the same as the previous release.
v0.2.0
- Introduced an interactive wizard for configuring agent-to-agent communication policies for OGP agents.
- Enables per-peer and global response policy configuration, including topic selection and response levels (`full`, `summary`, `escalate`, `deny`).
- Provides detailed CLI usage examples for viewing, setting, and managing policies.
- Documents inheritance, storage schema, and real-world configuration scenarios.
- Adds activity logging and troubleshooting guidance for policy issues.
Metadata
Frequently Asked Questions
What is ogp-agent-comms?
Interactive wizard to configure agent-to-agent communication policies (updated for multi-framework `--for` workflows, OGP 0.2.24+ peer identity, and 0.2.28+... It is an AI Agent Skill for Claude Code / OpenClaw, with 213 downloads so far.
How do I install ogp-agent-comms?
Run "/install ogp-agent-comms" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is ogp-agent-comms free?
Yes, ogp-agent-comms is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does ogp-agent-comms support?
ogp-agent-comms is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created ogp-agent-comms?
It is built and maintained by latentgenius (@dp-pcs); the current version is v0.6.0.
More Skills