CLI Agent Orchestrator is a tool that coordinates multiple AI coding command-line programs by running them as separate workers in isolated terminal sessions. A supervisor uses it to delegate software tasks to specialist agents in parallel or in sequence while the workers retain their normal command-line capabilities. The catalogue skills operate this orchestration workflow.
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 awslabs/cli-agent-orchestrator --skill cao-session-managementgit clone --depth 1 https://github.com/awslabs/cli-agent-orchestratorWrote 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/awslabs/cli-agent-orchestrator/cao-session-management)<a href="https://agentmods.dev/skills/awslabs/cli-agent-orchestrator/cao-session-management"><img src="https://agentmods.dev/badge/skills/awslabs/cli-agent-orchestrator/cao-session-management/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/awslabs/cli-agent-orchestrator/cao-session-management"><img src="https://agentmods.dev/badge/skills/awslabs/cli-agent-orchestrator/cao-session-management.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 7 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Data Exfiltration · line 112 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 113 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 114 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 115 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 131 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 151 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
- high Data Exfiltration · line 165 Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.Fix: Remove any code that sends prompts, responses, or session data externally. Preserve user privacy; never exfiltrate conversation content.
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 | $0.00083 | $0.01314 |
| Opus 5.5 | $0.00033 | $0.00526 |
| Sonnet 5.5 | $0.00017 | $0.00263 |
| Haiku 4.5 | $0.00008 | $0.00131 |
Grade B, and why
cao-session-management scanned grade B 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 13d 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.
Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
tool restrictions, so the agent can run any command, including `aws`, `rm` and Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
`curl`. Use `--yolo` only when the user asks for it. How it starts
The opening of the file, as written. The whole thing — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CAO Session Management
Operating CAO
Run cao --skill for the CAO operating guide: preflight, launch, session commands,
statuses, delegation, and safety. Run cao <group> --help for exact syntax.
Two rules apply to every launch from an agent:
- Use
--auto-approve. It skips the confirmation prompt and leaves the profile's tool policy unchanged; whether that policy is enforced depends on the provider (read theEnforcement:line launch prints).--yoloalso skips the prompt but removes all tool restrictions, so the agent can run any command, includingaws,rmandcurl. Use--yoloonly when the user asks for it. - If a
caocommand cannot reach the server, check it withcurl -sf http://localhost:9889/health.
A reported status is inferred from the rendered terminal screen, so it can disagree with reality. Before reporting readiness, progress, or completion to a user, corroborate the status with an output read; see cao-session-liveness.
Discovering Available Profiles
Profiles are CAO-level entities, installed with cao install regardless of which CLI
provider runs them. To find available profiles:
| Source | Command |
|---|---|
| All available profiles, with the source each one resolved from | cao profile list (no server needed), or curl -sf http://localhost:9889/agents/profiles |
| Profiles matching a keyword | cao profile find "<keywords>" |
| Custom/local profile files only | ls ~/.aws/cli-agent-orchestrator/agent-store/ |
Profiles installed via cao install <name> |
ls ~/.aws/cli-agent-orchestrator/agent-context/ |
| Profile installation and keyword discovery | see Agent profile installation and profile discovery |
Provider-native list (kiro_cli only) |
kiro-cli agent list, useful because CAO mirrors profiles into ~/.kiro/agents/ |
cao profile list and the HTTP endpoint return the same list. Discovery scans the
local store (agent-store/) first, then provider-specific directories (including
agent-context/), then extra directories from settings, and the built-in packaged
store last. The first match wins, so a local copy of a profile shadows the built-in
one with the same name. Each entry has a source label showing where it came from.
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.
- 13d ago Changed · -57 lines · +34 tokens per session scan A → B 57d5d60f5e15
- 1mo ago First seen · 168 lines · 49 tokens per session scan A 9812c0c8e23f
cao-session-management is a skill published in the GitHub repository awslabs/cli-agent-orchestrator (1,465 stars, last pushed today), licensed Apache-2.0. It adds 83 tokens to every session and 1,314 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it B with 2 findings (unrestricted tool access, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
agent-self-scheduling
Schedule AI agent runs with cron, loops, or external clocks while avoiding unsafe tight autonomous timers.
loop
Run a prompt or skill in this session on a recurring or variable interval (e.g. /loop 5m /foo).
dev-cleanup
Bounded checklist for cleaning local dev artifacts (build outputs, caches, temporary logs) in a single repository. NEVER auto-executes destructive commands — every cleanup action requires explicit, in-the-moment user confirmation. Use when a user explicitly asks for cleanup, or when the orchestrator determines stale…
tool-selection
Internal guidance for choosing between agent-rack's sync and background execution tools, and when to use per-agent shortcuts. Use whenever delegating a task to a claude/codex/opencode/Antigravity/custom sub-agent through agent-rack.
review
Run a structured, read-only code review through agent-rack's agentreview tool.
session-send
Send follow-up input to a running background agent-rack session.