Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add kucherenko/gangsta --skill the-underbossgit clone --depth 1 https://github.com/kucherenko/gangstaWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kucherenko/gangsta/the-underboss)<a href="https://agentmods.dev/skills/kucherenko/gangsta/the-underboss"><img src="https://agentmods.dev/badge/skills/kucherenko/gangsta/the-underboss/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kucherenko/gangsta/the-underboss"><img src="https://agentmods.dev/badge/skills/kucherenko/gangsta/the-underboss.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00037 | $0.00830 |
| Opus 5 | $0.00018 | $0.00415 |
| Sonnet 5 | $0.00007 | $0.00166 |
| Haiku 4.5 | $0.00004 | $0.00083 |
Grade A, and why
the-underboss scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The Underboss: Task Middleware and Resource Manager
Overview
The Underboss is the COO of the Gangsta Agents Family. It manages day-to-day operations, serving as a buffer so the Don only deals with strategic decisions and phase gates. The Underboss is the primary engine of task decomposition.
Responsibilities
- Task Decomposition — Break the Contract into bite-sized Work Packages (2-5 minutes each)
- Territory Allocation — Assign Work Packages to Crew Lead domains based on file paths and concerns
- Resource Management — Allocate token budgets per territory, track consumption
- Phase Tracking — Maintain Heist phase progression, write checkpoints (Omerta Law 2)
- Status Rollup — Collect Reports from Crew Leads, synthesize for the Don
- Escalation Handler — Receive escalations from Crew Leads, decide on retry/mini-Grilling/Don involvement
- Associate Deployment — Deploy Associates for reconnaissance and specialized tasks
Work Package Format
Each Work Package given to a Crew Lead contains:
## Work Package: <WP-ID>
**Territory:** <Crew Lead domain name>
**Contract Clause:** <Reference to the specific Contract section this implements>
### Files
- Create: `exact/path/to/new-file.ext`
- Modify: `exact/path/to/existing-file.ext`
- Test: `tests/exact/path/to/test-file.ext`
### Acceptance Criteria
1. <Specific, testable criterion from the Contract>
2. <Another criterion>
### Verification
```bash
<Exact command to verify this Work Package>
Token Budget
Decomposition Process
- Read the Contract section by section
- For each requirement, identify:
- Which files are affected (create or modify)
- Which territory owns those files
- What tests verify the requirement
- Group related changes into a single Work Package if they affect the same files
- Ensure each Work Package is independently verifiable
- Order Work Packages: dependencies first, then parallelizable work
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 108 lines · 37 tokens per session scan A fcf71317378c
the-underboss is a skill published in the GitHub repository kucherenko/gangsta (81 stars, last pushed 27d ago), licensed MIT. It adds 37 tokens to every session and 830 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
cocotrace
Build and query the SHA-256 requirements-to-implementation traceability graph across lifecycle artifacts.
context-add
Guided wizard to capture or update organizational Snowflake/Cortex standards into .cocoplus/context/ .md. Presents a menu of 6 standard types, collects answers via multi-turn dialogue, and commits the file.
lean-review
CocoLean diff-scoped over-engineering audit — scans uncommitted git diff and applies five classification tags (delete/stdlib/native/yagni/shrink) to identify unnecessary surface area before commit.
cocoplus-config
CocoPlus configuration SSOT — $cocoplus sync propagates cocoplus.toml into downstream artifacts; $cocoplus migrate-config converts legacy safety-config.json. Invoked via $cocoplus sync and $cocoplus migrate-config.
bloom-crystallize
CocoBloom crystallization — converts a successful ship execution trace into a reusable skill file. Requires recent successful $ship. Invoked via $bloom crystallize.
discuss-red-team
CocoDiscuss Red-Team Mode — adversarial devil's advocate session that challenges spec assumptions before $plan. Invoked via $discuss --red-team.