Sweep the full current session, repo/task context, local task files, and (with approval) connected remote task stores for outstanding actionable work, then record only useful high-confidence items in docs/outstanding-issues.md with priority, context, and next steps. Use when the user types "/ledger" or asks to extract outstanding tasks from this session into the ledger.
89%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required sweep step 4 (SKILL.md lines 27-29) reads GitHub PRs/issues and Todoist or other connected MCP task stores, where outsiders can post text that the workflow ingests; per the GitHub/Jira rule, PR and issue reading is medium even though the read is gated by user approval.
BigSimmo/PsychSift
fada614
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.