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/keefar/ticket-flow/flownpx skills add keefar/ticket-flow --skill flowgit clone --depth 1 https://github.com/keefar/ticket-flowWhat 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 | $0.00292 | $0.10701 |
| Opus 5 | $0.00146 | $0.05351 |
| Sonnet 5 | $0.00058 | $0.02140 |
| Haiku 4.5 | $0.00029 | $0.01070 |
Grade A, and why
flow 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 2d 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 — 402 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/flow — Ticket-Flow orchestrator
Args:
<kanban-id>(required — except with--parallel) ·<branch-suffix>(optional, forwarded to /pickup) ·--parallel(optional — work multiple tickets at once via worktree-isolated subagents; with no id = the whole ready queue, see ## Parallel mode) ·--local(optional, the default — kept as explicit flag for symmetry) ·--serial(optional — modifier of the subagent machinery, implies--parallel: one subagent at a time, merge + deploy + cleanup per ticket right after it returns; see ## Serial and loop) ·--loop(optional — modifier, implies--parallel: after every merge re-query the ready queue and continue until it is empty; takes no ids) ·--here(optional, local mode only — forwarded to/ticket-flow:pickup --here: adopt the worktree/branch this session is already in — orca, Conductor, worktrunk, bead-workflow-skills — instead of creating one; rejected with--parallel/--serial/--loop) ·--decisions a,b,c/--use-recommendations(optional, mutually exclusive — resolve the spec's## Decisionssection; see step 1.7;--decisionsis rejected with--parallel/--serial/--loop).
Decide, don't prompt — clear-cut points get a default (all modes)
At any decision point that is unambiguous or indifferent, decide with a sensible default and proceed — never stop with an AskUserQuestion menu for it. Stop only for genuine ambiguity: a choice where a wrong pick has real, non-obvious consequences and no default is clearly right. This is unconditional default behavior across every mode (--local, --parallel) — not a flag, not a mode.
Three things still stop on purpose — they are neither clear-cut nor indifferent:
--local's per-phase checkpoints — the "Ready for /implement? / finish?" review gates in steps 3/4/5. These are the value of--local; they stay. A checkpoint-free local run, if ever wanted, is a separate future flag.--parallel's P5 consolidated checkpoint — merge all / subset / stop. Not in--serial: that modifier exists for unattended runs, so each ticket merges right after its subagent returns and P5 becomes a per-ticket report, not a question — the gates are finish's verification and the merge guard (finish Step 7). Want checkpoints? Use plain--parallel(one consolidated) or--local(per phase).- The step-1.7 decision gate — a spec's genuine
## Decisionssection is ambiguous by construction; resolved by the user or by--use-recommendations, never auto-decided here.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago First seen · 402 lines · 0 tokens per session scan A d3f8ceeb8002
flow is a skill published in the GitHub repository keefar/ticket-flow (4 stars, last pushed 6d ago), licensed MIT. It adds 292 tokens to every session and 10,701 once invoked, about $0.0015 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 2026-08-31.
Other skills, from other repositories
kanvibe-release-deploy
Use this skill whenever releasing or deploying KanVibe desktop from a clean, up-to-date dev checkout: ask only for the target version and release-note approval, then let the AI update package versions, run pnpm run deploy, publish the DMG GitHub release, update the Homebrew cask, create the release PR, and auto-merge…
council-router
Route code reviews to appropriate council members. Use when reviewing PRs, architecture decisions, or significant code changes that need expert perspective.
issue-triage
Triage Paperclip inbox issues that are stale, blocked, in-review, or assigned-but-not-progressing, and decide a single next action per issue (resume, reassign, unblock, escalate, or close).
taiyi-ultrawork
TaiyiForge ultrawork — 并行切片 + Cursor Task 自动派发契约（对标 OMC ultrawork + spawnagent）.
codekanban-cli
Operate CodeKanban workflows, terminal sessions, and web sessions through the installable codekanban-cli command. Use when the user wants to create, inspect, control, watch, or continue CodeKanban AI work from a project path, project ID, or project name without relying on repository-local helper scripts. Prefer…
triage
Work-unit triage for GitHub issues. Groups raw issues, fuses each group with the AGENTS.md northstar, and externalizes each routed unit to a substrate record a collaborator session is pointed at.