The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to utilize standard shell utilities (grep, sort, tail, sed) to manage its observation log. These commands are used to compute unique identifiers for new observations, prevent numbering collisions, and manage log archival through check-then-act logic.
[EXTERNAL_DOWNLOADS]: The skill fetches documentation and user guidance from the official Rebelytics GitHub repository (github.com/rebelytics/one-skill-to-rule-them-all) when web access is available. These downloads are used to provide the agent with current methodology and licensing instructions.
[DATA_EXFILTRATION]: While the skill processes session history and user feedback, it employs a rigorous five-layer confidentiality framework to scrub sensitive information. This protocol includes observation-level stripping, pre-creation review, post-draft sweeps, and cross-product re-identifiability checks to ensure that no client-identifying or proprietary data is included in generated 'open-source' skills.
[SAFE]: The skill operates strictly within the user's workspace folder, staging updates for manual review and installation by the user. It does not attempt to modify system files, canonical skill mounts, or execute code from untrusted remote sources. All data processing is protected by documented privacy safeguards.

task-observer