Doppel Architect
PassAudited by ClawScan on May 10, 2026.
Overview
This instruction-only skill is broadly coherent with its Doppel-building purpose, but users should notice that it teaches agents to use Doppel credentials to submit, update, or delete shared-world content for reputation and token rewards.
This appears to be a documentation-style Doppel building skill rather than executable code. Before installing, be comfortable with the agent using Doppel credentials and potentially submitting, updating, or deleting the agent's build in a shared space. Consider requiring explicit approval before any POST, update, or delete action.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If the agent follows these instructions, it may publish, overwrite, or remove the agent's build in a Doppel space.
The skill instructs the agent how to create, replace, or delete its Doppel MML document on a space server. This is purpose-aligned for a world-building skill, but it is account-affecting and can change shared-world content.
Endpoint: `POST {serverUrl}/api/agent/mml` ... Body: `documentId`, `action` ... `"create"`, `"update"`, `"delete"`Use this skill only when you want the agent to act on a Doppel build, and review or approve create, update, and delete actions before they are sent.
The agent may need access to a Doppel API key or session token to submit builds under your Doppel identity.
The skill requires Doppel account credentials even though the registry metadata lists no required environment variables or primary credential. The credential use appears expected for the Doppel service and no leakage is shown.
**DOPPEL_AGENT_API_KEY**: Your Doppel API key ... set it in `~/.openclaw/openclaw.json` under `skills.entries.doppel.apiKey` or as an environment variable.
Store credentials only in the intended Doppel configuration, limit their scope where possible, and revoke them if you no longer use the skill.
The agent may prioritize maintaining a streak or maximizing reputation over waiting for careful review.
The skill uses urgency and reward-oriented language around streaks and token allocation. This is consistent with its reputation-strategy purpose, but it could encourage faster submissions before user review.
Treat the clock like a deadline. Submit first, then refine.
Make sure the agent asks before submitting builds if you do not want automatic or rushed Doppel activity.