Execute one-shot or batch tasks using provider subagents.
npx skills add https://github.com/ethanhq/cc-fleet --skill subagent-ethanhq
Please help me install this Agent Skill. Skill: subagent Source: https://github.com/ethanhq/cc-fleet/tree/main/codex-plugin/skills/subagent Command: npx skills add https://github.com/ethanhq/cc-fleet --skill subagent-ethanhq
This Skill allows you to offload one-time tasks to provider subagents, utilizing different models and services, without the need for long-term collaboration.
Run a subagent to analyze the complexity of quicksort using the 'deepseek' provider:
ccf subagent deepseek --model strong \
--prompt "Analyze the worst-case complexity of quicksort in src/sort.go; give a triggering input" \
--label "codex/quicksort" --json
High-intent search queries and answers about installing and using this skill.
You can execute parallel research and bulk processing tasks by offloading them to provider subagents, which supports batch fan-out across various models and services. This enables independent tasks to run simultaneously without long-term collaboration.
Yes, provider subagents are designed for one-shot tasks like code analysis, research, or judgment. You can execute a single task using a specific model, such as deepseek, and receive the output directly without maintaining a persistent agent.
To offload tasks to provider subagents, you must configure your provider API keys and complete the Claude Code setup. This environment preparation ensures the subagents can authenticate and communicate with the desired models.
You can control resource usage during batch processing by utilizing cost-bounded probes. This feature limits the duration and cost of subagent tasks, ensuring your parallel model execution stays within defined resource constraints.
Yes, subagent execution supports provider flexibility, allowing you to utilize a wide range of models from various providers. This includes using Codex or Claude subscriptions to handle specialized model tasks for one-off execution.
alibabaConfigure Higress AI Gateway as a model provider for Clawdbot/OpenClaw.
profmikegreeneParses pilot plans and hypotheses into structured JSON or tabular data for downstream analysis.
codekilnAutomate batch replies to GitHub PR review comments via gh api.
dangdungvnOverride Riverpod providers with overrideWith, overrideWithValue, and overrideWithBuild.