Menu
Batch study all repos/tools in .references/. Dispatches parallel subagents to document each, then distills learnings into steering/skills. Use when onboarding to a project with multiple reference repos.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Dispatch Codex to review all unreviewed repository work, then verify its correlated findings ticket or clean result. Use when requesting an independent Codex review, auditing all changes, checking work since the last review, or waiting for Codex findings. Trigger: dispatch codex review, independent review, review all changes, check codex findings, run review agent.
Always-on: when a project has tickets, work the frontier (first unblocked open ticket), mark done, update plan status.
Incremental and full-history repository review with resumable adoption batches. Use when reviewing new work, initializing review tracking, reviewing all work to date, resuming historical review, inspecting unreviewed commits, or reviewing changed tickets. Trigger: review new work, adopt repository, review all history, resume review, review marker, unreviewed commits.
End-of-session handoff that captures current state for the next session. Use when ending a work session, switching context, or before a long break.
Consolidate project artifacts — promote scratch to memory, deduplicate memory, process decisions, organize scripts, update steering/skills accuracy. Use periodically or when the project feels cluttered.
Run and interpret crew-research evals — run.sh flags, scores.jsonl fields, activation TPR/FPR verdicts, resuming interrupted runs. Use when running evals, reading eval results, diagnosing a failing definition, or resuming a dead run. Trigger: run the evals, eval results, scores.jsonl, activation test, TPR, delta threshold, resume the run, skip-completed, known gap.
| name | study-all-references |
| description | Batch study all repos/tools in .references/. Dispatches parallel subagents to document each, then distills learnings into steering/skills. Use when onboarding to a project with multiple reference repos. |
| metadata | {"type":"process","invocation":"user-only","practice":null} |
Analyze every repo/tool in .references/ and extract actionable knowledge.
List all directories in .references/. For each, note what it appears to be (language, purpose, size).
For each reference, check if .memory/{name}-reference.md already exists with studied_at frontmatter. Skip those — already studied. Only dispatch subagents for unstudied references.
Dispatch one subagent per unstudied reference. Each doc follows the study-reference skill's schema — same frontmatter (studied_at, source: .references/{name}), same seven analysis dimensions (JTBD, Decisions and tradeoffs, Non-obvious, Prior art, Audience, Conventions, Integration points). This skill is the batch driver; study-reference owns the document structure.
Subagent prompt template:
Explore .references/{name}/ thoroughly. Read key files (README, main entry points, config).
Write a reference doc to .memory/{name}-reference.md following the structure in
.kiro/skills/study-reference/references/exploration-template.md (JTBD, Decisions and
tradeoffs, Non-obvious, Prior art, Audience, Conventions, Integration points), with
frontmatter: studied_at ({ISO 8601 timestamp}) and source (.references/{name}).
Focus on what's relevant to THIS project (read AGENTS.md for context).
After all subagents complete, read all .memory/*-reference.md files and produce:
Patterns that appear in multiple references (consensus = high confidence).
Rules that should apply every turn. For each:
On-demand knowledge for specific tasks. For each:
New vocabulary from the references.
Present all candidates. After user approves:
.kiro/steering/.kiro/skills/.memory/CONTEXT.mdAGENTS.md with new tool referencesdocs: study all references — {summary of learnings}.memory/{name}-reference.md (already studied).