worktrail-help

Show the Worktrail front-door commands and accepted positional arguments. Use when the user asks how to invoke worktrail-go, wants the available handoff, spec, pr, repo, or route forms, invokes worktrail-go with help or with a bare noun, or asks how to answer a filed product decision (worktrail-decision).

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Worktrail command help

Display this concise command reference. Do not start work, claim a brief, or run the dashboard merely because help was requested.

Client command names

Use the command exposed by the host client:

Client	Front door
Claude Code	`/worktrail-go`
Codex	`$worktrail:worktrail-go`

The help skill itself is available as /worktrail-help in Claude Code or $worktrail:worktrail-help in Codex.

Accepted forms

This block is generated from the parser's own form registry (worktrail-go-parse --forms) and pinned by tests/router/test_parse_invocation.py, so it cannot drift from what the front door actually accepts. Regenerate it rather than editing it by hand.

<front-door>                                     dashboard and interactive picker
<front-door> help [topic]                        show this reference
<front-door> <repo>                              show active work for a repository
<front-door> handoff new <focus>                 capture a new handoff brief
<front-door> handoff list                        list queued handoff briefs
<front-door> handoff start <id>                  claim or resume a queued handoff
<front-door> handoff auto                        auto-pick one ranked queue brief
<front-door> handoff drain [max-items] [repo]    drain multiple items in fresh contexts
<front-door> spec new <request>                  plan a new feature
<front-door> spec implement <id>                 execute a specification
<front-door> spec continue [<id>]                resume in-flight work
<front-door> spec fix <request>                  defect repair (Route F)
<front-door> spec explore <idea>                 idea discovery (Route A)
<front-door> spec route <A-J> [<id>]             force a route
<front-door> pr fix                              PR / CI repair
<front-door> decision list                       list open product decisions awaiting an answer
<front-door> decision answer <decision-id>       answer a filed decision interactively
<front-door> <brief-id>                          claim or resume a queued handoff (same as handoff start)
<front-door> <free text>                       * classify and route any other request

* not a parsed form -- reaches the route classifier as free text
Every <noun> <verb> form takes an optional leading <repo> to scope it, e.g. <front-door> <repo> spec implement <id>.

Compatibility spellings, still accepted and read as the form shown:
<front-door> auto                      = <front-door> handoff auto
<front-door> drain [max-items] [repo]  = <front-door> handoff drain [max-items] [repo]
<front-door> handoff:<id>              = <front-door> handoff start <id>
<front-door> new <request>             = <front-door> spec new <request>
<front-door> implement [spec] <id>     = <front-door> spec implement <id>
<front-door> continue [<id>]           = <front-door> spec continue [<id>]
<front-door> fix <request>             = <front-door> spec fix <request>
<front-door> brainstorm <idea>         = <front-door> spec explore <idea>
<front-door> spec brainstorm <idea>    = <front-door> spec explore <idea>
<front-door> route:<A-J> [<id>]        = <front-door> spec route <A-J> [<id>]
<front-door> pr                        = <front-door> pr fix

The grammar is <front-door> [<repo>] <noun> <verb> [args] with four nouns: handoff (the work queue), spec (spec-driven work), pr (an open pull request), and decision (the human-decision queue). Forms are listed in the order the parser checks them, so an earlier line wins over a later one. A noun typed without a recognised verb shows this reference for that noun instead of being routed as free text.

Two bare shortcuts sit outside the noun-verb shape on purpose: <brief-id> (the same as handoff start <id>) and <repo>. <brief-id> is a Worktrail queue handoff identifier, such as 20260726-140500. A specification identifier is deterministic when used with spec implement <id> or spec route <A-J> <id>. A bare spec name may instead be interpreted through the dashboard picker.

handoff auto runs one queue item. handoff drain runs multiple items with fresh contexts. The standalone drain skill is intentionally not user-facing.

The compatibility spellings are permanent, not deprecated: nothing that worked before the noun-verb grammar stops working. handoff:<id> in particular stays the internal dispatch spelling between worktrail-go and its executor; the front-door spelling is handoff start <id> or the bare <brief-id>.

Answering a filed decision

An unattended run that hits a genuine product decision files a record instead of guessing, then blocks the brief until you answer. <front-door> decision list shows what is open, and <front-door> decision answer <decision-id> walks you through that record's Question/Background/Options interactively and records your answer. The bare <front-door> dashboard surfaces the same open decisions as a picker category. Full interactive procedure: ../worktrail-go/references/answer-decision.md.

For non-interactive use (e.g. scripting an answer), the underlying CLI is still available directly:

worktrail-decision list                         everything open/answered/resolved
worktrail-decision answer <id> --answer "..."    record your answer

The blocked brief unblocks automatically once you answer; the next drain pass picks it back up. Full filing/resuming procedure: ../worktrail-go/references/decision-queue.md.

Repository: behindthedash/worktrail
Path: skills/worktrail-help/SKILL.md
Commit: ca98247

Last updated: ~2 hours ago
First committed: 11 days ago

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.