@hollis9087

hermes-continuation-maintainer diagnoses and repairs premature conversation loop terminations in hermes agent runs by scanning logs for misclassified stops, applying targeted heuristic patches to control flow, and validating via regression tests.

agent-memory-cleanup detects and remediates degraded agent memory files (user.md, memory.md, profile.md, etc) via staged audit, proposal, and approval flows. triggers on memory pressure signals or explicit user request. creates timestamped backups before applying changes.

holli-financial-fraud-analyzer applies beneish m-score and dechow f-score models to detect earnings manipulation and accounting anomalies from multi-period financial statements. outputs structured risk assessment with three-layer evidence integration and industry-adjusted thresholds for tech, manufacturing, and consumer sectors.

agent-workspace-hygiene inventories and optionally archives generated build, cache, and test artifacts from coding repos. it requires explicit user request and defaults to dry-run, recoverable moves over deletion.

holli-blogwatcher wraps a go CLI for tracking RSS/Atom feeds and blog updates. covers add/list/scan/read operations but lacks integration guidance and failure handling.

holli-academic-reader applies control theory frameworks to technical book analysis, offering chapter summaries and reflection prompts. core implementation details and execution mechanics are largely absent.

Use this skill only when the user explicitly asks for a code audit, security audit, risk-focused PR/diff review, repo or module audit, regression-risk review...

Ultra-light automatic handoff manager for long-running agent sessions at context-boundary moments. Use this skill when a runtime reports context compaction/s...

Audit and operate the Trinity/OpenClaw evolution loop by validating status, preflight, capability gates, repair fallback, and user progress reports.