# Lissom Research > Dispatches lissom-specs-reviewer and then lissom-researcher agents optionally to refine the specification and produce the research document given an explicit task_dir, which contains the original specification. Use when this capability is needed. - Skill: `tomevault-io/lissom-research` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/lissom-research` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/lissom-research/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/lissom-research --- ## Inputs - `task_dir` = "$0" - `user_attention` = "$1" (optional): `default` (default), `auto`, or `focused` - `spec_review_required` = "$2" (optional): `no` (default) or `yes` - `research_required` = "$3" (optional): `yes` (default) or `no` ## Process Execute sequentially: 1. **Conditional**: Skip this step if `user_attention` is `auto` OR if `spec_review_required` is `no`. Otherwise: Use Tool `Agent` to spawn `lissom-specs-reviewer`, passing it the `task_dir` and `user_attention`. - If it returns `Specs INCOMPLETE` (auto mode only), relay the reasons to the user as a warning, then proceed. - Any other return value: treat as success and proceed. - If `/Specs.md` does not exist or is empty after the agent returns, escalate immediately. 2. **Conditional**: Skip this step if `research_required` is `no`. Otherwise: Use Tool `Agent` to spawn `lissom-researcher`, passing it the `task_dir` and `user_attention`. ## Completion Return to the caller only after `/Research.md` exists and is non-empty. If it does not exist, re-invoke `lissom-researcher` once before escalating. --- > Source: [cuzfrog/lissom-skills](https://github.com/cuzfrog/lissom-skills) — distributed by [TomeVault](https://tomevault.io).