Elen: Epistemic Decision Tracker
PassAudited by ClawScan on May 1, 2026.
Overview
The skill is coherent and purpose-aligned, but users should notice that it relies on an external local MCP/npm server and creates persistent decision records that can influence future agent behavior.
This skill appears safe for its stated purpose if you want an agent to use Elen as a persistent decision log. Before installing, review the referenced @learningnodes/elen-mcp package, understand that it will run locally through MCP, and avoid storing secrets or sensitive details in the decision graph.
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.
When this skill is active, the agent may add decision-network checks and commits to development tasks.
The skill makes use of the MCP server a mandatory part of the agent’s decision workflow. This is consistent with the stated purpose, but it can alter the agent’s normal task flow.
You **MUST** use the Elen MCP server to document your epistemic decisions.
Use the skill only where persistent decision tracking is desired, and override or disable it for tasks where that workflow is not appropriate.
Installing the integration would run local code from the referenced npm package as part of the MCP server setup.
The skill instructs configuration of a local MCP server by running an external npm package. The package version is pinned and the behavior is purpose-aligned, but the package code is not part of the provided skill artifacts.
"command": "npx", "args": ["-y", "@learningnodes/elen-mcp@0.1.1"]
Review the referenced package and its publisher before enabling it, keep the version pinned, and install it only in environments where a local decision-graph server is acceptable.
Stored decisions may guide future technical choices, and sensitive project or security details could persist if committed.
The skill intentionally creates persistent decision context that can be searched and reused later. This is the core feature, but incorrect or sensitive entries could influence future agent work.
choices are version-controlled, preserved, and searchable
Avoid storing secrets, review important decisions before committing them, and use supersede or cleanup workflows for stale or incorrect entries.