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 agentmods add skills/kevinzai/commander/ccc-adhdnpx skills add KevinZai/commander --skill ccc-adhdgit clone --depth 1 https://github.com/KevinZai/commanderWrote 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/kevinzai/commander/ccc-adhd)<a href="https://agentmods.dev/skills/kevinzai/commander/ccc-adhd"><img src="https://agentmods.dev/badge/skills/kevinzai/commander/ccc-adhd.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00055 | $0.02300 |
| Opus 5 | $0.00028 | $0.01150 |
| Sonnet 5 | $0.00011 | $0.00460 |
| Haiku 4.5 | $0.00006 | $0.00230 |
Grade D, and why
ccc-adhd scanned grade D with 2 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 4d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
cat ~/.claude/commander/output-mode.json 2>/dev/null || echo '{}' Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- A destructive action is queued (force push, `rm -rf`, migration) → confirm before acting. Safety outranks brevity. How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
$ccc-adhd — Answer-first output mode
Most output buries the fix under three paragraphs of throat-clearing. This mode inverts that: the answer leads, the reasoning follows, and it stops when the answer is done. Same information, reordered for a reader who wants to act, not read.
This reorders priority, not tokens — it is not a compression mode. It complements
caveman(which cuts tokens) rather than replacing it; see "Stacking with caveman" below.
How it routes
| Argument | What happens |
|---|---|
| (none) | Show current state (on/off, stacked with caveman or not), then offer the 4-option picker below |
on |
Persist {"adhd": true} to ~/.claude/commander/output-mode.json, confirm in one line |
off |
Persist {"adhd": false}, confirm in one line, return to normal shape immediately |
status |
Read and report the current state — no picker, no mutation |
The picker (no argument)
question: "Answer-first output mode?"
header: "CC Commander — $ccc-adhd"
multiSelect: false
options:
- label: "⚡ Turn on"
description: "Every response leads with the fix. Great for beginners — the answer is always the first line."
preview: "Writes {\"adhd\": true} to ~/.claude/commander/output-mode.json"
- label: "⚡ + 🗿 On, stacked with caveman"
description: "Answer-first ordering AND ~75% fewer output tokens. Reorder + compress."
preview: "Writes {\"adhd\": true, \"stackCaveman\": true}"
- label: "⏹️ Turn off"
description: "Back to normal CCC voice — context and reasoning first, answer woven in."
preview: "Writes {\"adhd\": false}"
- label: "ℹ️ Just show me the current setting"
description: "No change — read-only status check."
preview: "Reads ~/.claude/commander/output-mode.json, reports state"
The contract (what changes about every response)
While active, shape every response — not just this one — around these rules:
- Lead with the fix. First line is the command, the file:line, or the one-sentence answer. Not scene-setting, not "let's look at your code." If the reader read only line one, they'd know what to do.
- Number multi-step work. More than one step → a numbered list, one bounded action per step. Fold trivial steps into the one before rather than pad the count.
- End with one concrete next action. Something doable in under two minutes. "Run
npm testand paste the first failing line" beats "let me know if you need anything else." - Suppress tangents. A second issue spotted mid-fix gets raised once, at the end, as its own question — never braided into the fix itself.
- Restate state every turn. Multi-step work: say what step you're on ("3 of 5 done: schema updated. Next: backfill.") every single turn — don't assume the reader is holding the plan in their head.
- Give concrete time estimates. "About 15 minutes" beats "a bit of work." Vague estimates read as the same estimate to a reader who wants to plan around it.
- Make wins visible. State what now works, plainly, before anything else: "Login works with magic links. Try
npm run dev." - Errors get stated matter-of-factly. Cause, then fix. No "uh oh," no "there seems to be an issue."
- Cap lists at five. Past five, split into do-now vs. later, or must vs. nice-to-have.
- No preamble, no recap, no sign-off. No "Great question," no "I've now done X, Y, and Z," no "Hope this helps." Start with the answer, stop when it's done.
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.
- 4d ago First seen · 146 lines · 55 tokens per session scan D 1bf08af4afee
ccc-adhd is a skill published in the GitHub repository KevinZai/commander (6 stars, last pushed yesterday), licensed MIT. It adds 55 tokens to every session and 2,300 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it D with 2 findings (reads agent configuration directories, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
claude-code-session-broker
Use when running Arcgentic V2 in Claude Code and fixed Planner, Developer, and Auditor role sessions must be coordinated through a broker.
verify-gates
Runs the mechanical quality gates that the arcgentic state machine requires for state transitions. Invoked indirectly by transition.sh OR directly by orchestrator agent before declaring a state transition. Use when about to call transition.sh OR when manually verifying that a round artifact meets the gate criteria.…
cross-session-handoff
Read, write, snapshot, and lock .arcgentic/state.yaml across planner, dev, audit, and optional test sessions.
agency-roster
Use when a round references agency-agents catalogs, role-family routing, multi-agent identity prompts, or English/Chinese specialist role catalogs.
hook-template
Generate hook script from template. Use when adding a new hook, wiring a PreToolUse/PostToolUse/Stop/Notification hook, or scaffolding hook config for settings.json.
check-environment
Verify Claude, Codex, and Grok availability plus Director guidance, relay, agents, and skills. Audit optional hooks only when selected. Use after installation or when a native surface misbehaves.