agent-workflow

Plugin agent-workflow from marketplace curtisg-skills in curtisgalloway/public-skills. Commands read from the repo's manifest — 8 days ago.

Skills (17)

• agent-agnostic-skills

  How to write skills, hooks, subagent definitions, and agent-facing scripts that survive a change of harness (Claude Code, Antigravity, whatever ships next), and how to port one that didn't. Use when authoring or reviewing any of those, before hardcoding a path, variable, or tool name another harness

• claude-model-selection

  Choose between Claude Opus 5.5 and Fable 5.1 for a session, subagent, or scheduled run, accounting for the billing mode (Pro, Max, Team or Enterprise seat, or API) and deciding when to escalate from one to the other. Use when the user asks "Opus or Fable?", "which Claude model should I use for this"

• claude-session-transcript

  Reference doc — how to locate the live Claude Code session transcript on disk and extract real user prompts from it (filter rules, slash-command pair collapsing, noise stripping). Other skills (learn, teach) Read this file before parsing transcripts. Not a user-invocable skill.

• codex-model-selection

  Choose GPT-6.1 Sol, GPT-6 Astra, or GPT-6 Luna for a Codex session, subagent, or scheduled run, accounting for subscription allowances, paid credits, API billing, reasoning effort, and escalation. Use when the user asks "which Codex model", "Sol or Astra", "can Luna handle this", or wants a Codex mo

• consult

  Collaborate with another coding agent on a question or proposed approach: any of Claude Code, Codex, Antigravity (agy) and Muse consults another, or a different model in its own CLI (Opus consulting Fable), through a persistent peer session. Use when the user says "consult with <name>", where the na

• design-partner

  Adopt a thinking-partner posture for design, architecture, and brainstorming: explore the problem, lay out options and tradeoffs, push back — without touching code. Use when the user wants to think something through ("let's brainstorm", "should I", "talk me through the tradeoffs") rather than have a

• gdoc-review-loop

  Review a repo-owned Markdown document with a stakeholder through Google Docs, in numbered rounds. The Markdown file in git stays the source of truth; each round is a new Doc built from it, the reviewer's direct edits and margin comments are read back and applied to the file, and the reply to their c

• git-delegate

  Save main-model tokens by handing git work to a subagent on a cheaper model in Claude Code or Codex through a fixed brief, and taking back a report of ten lines or fewer instead of raw logs and diffs. Covers inspection (log, diff, blame, cherry) and local writes (stage, commit, branch, rebase, prune

• handoff

  Preserve working state across a context clear or agent restart. Write mode dumps a HANDOFF.md the next session can cold-start from (task, state, decisions, dead ends, next steps); resume mode reads it back and continues the work. Use when the user invokes /handoff, says they're about to clear contex

• intern-mode

  A loop-safety posture: after 12 turns without meaningful progress, stop, file a stuck report, and wait for direction. Use when the user wants protection against runaway loops or silent thrashing ("intern mode", "stop if you get stuck"). Stays active until explicitly released.

• lab-notebook

  Keep a running lab notebook and a process log while working: one append-only, timestamped chapter per unit of work (a milestone, an investigation, a debugging session), a timestamped index that makes the chapters searchable, and a per-project log of where the agent's process cost time, as input for

• learn

  Two modes. Quick (default): review the current session transcript for lessons that would have made it go smoother — failed commands, wrong tool arguments, user corrections, environment surprises — plus the project's lab-notebook process log when one exists, and propose additions to the workspace or

• orchestrate-milestones

  Run a project's remaining implementation-plan milestones end to end as an orchestrator: recover state (including after a crash or interrupted session), then for each milestone in turn launch a fresh subagent that executes it by the project-plan process, verify its report, push its branch, open a pul

• project-plan

  Plan project work by creating or updating a detailed design document (approved by the user before planning), then deriving an implementation plan with cohesive, session-sized milestones and mandatory testing and review gates, then executing one milestone per session with recorded evidence and a hand

• quota-strategy

  Stretch Claude or Codex subscription allowances across long unattended runs by reading available usage pools, reserving capacity for the orchestrator, and routing implementation to a cheaper model or the other provider. Works with either orchestrator and with one or both providers configured. Also s

• run-delegate

  Save main-model tokens on commands whose output is long but whose answer is short — test suites, builds, lints, CI logs, server logs — by either trimming the output with quiet flags or handing the run to a subagent on a cheaper model in Claude Code or Codex that returns a capped report. Holds the sh

• teach

  Summarize past coding-agent sessions into teaching-oriented markdown files. Reads session transcripts from the agent's project-specific transcript directory, finds sessions that haven't been summarized yet (keyed on session_id in existing summary frontmatter), and writes one summary per unsummarized

Hooks (2)

• PreToolUse

  python3 "${CLAUDE_PLUGIN_ROOT}/skills/consult/scripts/runner_guard.py"

• PreToolUse

  python3 "${CLAUDE_PLUGIN_ROOT}/skills/orchestrate-milestones/scripts/implementer_guard.py"

Similar plugins