Superlazy Subagents
Silent subagent-driven plan execution. Use when executing an implementation plan with independent tasks and subagent support is available.
Install
$ agentstack add skill-starchybomb-superlazy-superlazy-subagents ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README — it links to the public security report.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-starchybomb-superlazy-superlazy-subagents) Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming — see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps — measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Superlazy Subagents
SUPERLAZY mode: output ≤3 lines (questions / user-actions / ✅❌); minimal code, full quality; do the work, never narrate it.
Loop (per task, no user check-ins between tasks)
- Extract the task's text to a brief FILE. Dispatch a fresh implementer subagent: one line of project context + brief path + interfaces/decisions from earlier tasks + report-file path. NEVER paste session history or prior-task summaries; never make it read the whole plan.
- Implementer questions → answer, re-dispatch. BLOCKED → more context, stronger model, or split the task — never blind retry.
- Review the task: generate diff as a FILE (
git diff BASE..HEAD, BASE recorded before dispatch — notHEAD~1), dispatch a reviewer with brief + report + diff paths + verbatim spec constraints. Never pre-judge findings ("don't flag X") in the prompt. - Critical/Important findings → one fix subagent with ALL findings, then re-review. Minor → ledger.
- Append to ledger file (
.superlazy/progress.md):Task N: complete (commits X..Y). Ledger survives compaction — trust it +git logover memory; NEVER re-dispatch a task it marks complete. - All tasks done → one final whole-branch review (merge-base diff file, most capable model). Findings → ONE fixer for the full list. Then
✅.
Cost rules
- Cheapest model that fits, stated explicitly per dispatch: transcription-from-plan/1-file fixes → cheapest; multi-file integration or prose-spec implementers and reviewers → mid-tier; architecture + final review → most capable. Turn count beats token price — too-cheap models take 2-3× turns.
- Artifacts move as files (brief, report, diff), never pasted text.
- Never dispatch implementers in parallel (conflicts). Never skip re-review after fixes.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: StarchyBomb
- Source: StarchyBomb/superlazy
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.