Beckmann Knowledge Graph × Self-Improving Agent
PassAudited by ClawScan on May 14, 2026.
Overview
This instruction-only orchestration skill appears coherent and asks before using the Beckmann reasoning flow, but it relies on two other skills and writes persistent learning notes.
Before installing, make sure you trust and have reviewed the two prerequisite skills. Use it for non-sensitive deep-reasoning tasks unless you are comfortable with summaries and insights being saved in `.learnings/` and potentially promoted into persistent agent instructions.
Findings (2)
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.
The combined behavior may change depending on the installed versions and contents of the required skills.
The skill's behavior depends on two external skills that are not included in the provided artifacts. This is disclosed and central to the purpose, but those dependencies determine much of the actual runtime behavior.
requires: "pskoett/self-improving-agent, matthiasbeckmann987-spec/beckmann-knowledge-graph"
Review and install trusted versions of the referenced Self-Improving Agent and Beckmann Knowledge Graph skills before relying on this combination.
Sensitive questions or strategic reasoning summaries may remain in local learning files and could influence later agent behavior if promoted.
The skill persists summaries and insights into local learning files and may suggest promotion into persistent agent context. This is purpose-aligned for a self-improving workflow, but persistent memory can affect future answers.
After every Beckmann analysis, add an entry to `.learnings/LEARNINGS.md` ... `Promote to CLAUDE.md if broadly applicable.`
Review `.learnings/LEARNINGS.md` and any `CLAUDE.md` promotions, and avoid using the skill for sensitive topics unless you are comfortable with local persistence.