Skip to content
Testing
Skill

/pdca-fast-track

Daniel-mode fast-track auto-approves Checkpoint 1-8 when Trust ≥ 80, fastTrack on, Design doc exists. Else L2 + manual gates. Triggers: pdca fast-track, skip checkpoints, auto approve

From plugin
bkit
58944 skills34 agents2 commands21 hooks
Install
$ npx -y skills add popup-studio-ai/bkit-claude-code --skill pdca-fast-track --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/pdca-fast-track

Context preview

The summary Claude sees to decide when to auto-load this skill.

Daniel-mode fast-track auto-approves Checkpoint 1-8 when Trust ≥ 80, fastTrack on, Design doc exists. Else L2 + manual gates. Triggers: pdca fast-track, skip checkpoints, auto approve

SKILL.md

pdca-fast-track.SKILL.md
name: pdca-fast-track
classification: workflow
classification-reason: PDCA flow modifier — bypasses Checkpoint 1-8 user gates
deprecation-risk: none
effort: low
description: |
  Daniel-mode fast-track auto-approves Checkpoint 1-8 when Trust ≥ 80, fastTrack on, Design doc exists. Else L2 + manual gates.
  Triggers: pdca fast-track, skip checkpoints, auto approve
argument-hint: "<feature>"
user-invocable: true
allowed-tools:
  - Read
  - Write
  - Glob
imports: []
next-skill: null
pdca-phase: null
task-template: "[Fast-Track] {feature}"

PDCA Fast Track — Daniel's Track

> v2.1.11 Sprint β FR-β5. Wraps `lib/control/fast-track.js`. Provides a > single auto-escalation gate for users who have already earned Trust > ≥ 80 and produced a Design doc — Checkpoints 1-8 default to > Recommended without prompting. Falls back to L2 + manual review the > moment any preconditions fails (E-β5-01 / E-β5-02).

Arguments

| Argument | Description | Example | |----------|-------------|---------| | `<feature>` | Feature name to engage fast-track for | `/pdca-fast-track bkit-v2111-integrated-enhancement` |

Behavior

Preconditions (all must pass; fail-open)

1. **Config enabled** — `bkit.config.json#control.fastTrack.enabled === true` 2. **Trust Score ≥ 80** — `.bkit/state/trust-profile.json#trustScore` ≥ `fastTrack.minTrustScore` (default 80) 3. **Design doc exists** — resolved via `bkit.config.json#pdca.docPaths.design` templates against `<feature>`

When any block, surface the reason with a remedy hint:

| Block | Remedy hint | |-------|-------------| | `disabled_in_config` | "Set `control.fastTrack.enabled = true` in bkit.config.json" | | `trust_score_too_low` | "Run a few `/pdca check` cycles to raise trust" | | `design_doc_missing` | "Run `/pdca design <feature>` first" |

On engage (all preconditions pass)

1. Recommend Automation Level escalation L2 → `fastTrack.autoLevel` (default `L3`). 2. Persist engagement record to `.bkit/runtime/fast-track-log.json`:

   {
     "engaged": true,
     "feature": "<feature>",
     "engagedAt": "<ISO>",
     "trustScore": 85,
     "autoLevel": "L3",
     "fallbackLevel": "L2",
     "skipCheckpoints": [1,2,3,4,5,6,7,8],
     "designPath": "docs/02-design/features/<feature>.design.md",
     "decisions": []
   }

3. Each Checkpoint 1-8 invocation calls `fast-track.recordCheckpoint` to append a decision entry (audit trail). 4. If a Checkpoint surfaces NO "Recommended" option, call `recordCheckpoint({ decision: 'fallback', reason: 'no_recommended' })` and pause to `fallbackLevel`.

Security

  • All file IO is rooted at `process.env.CLAUDE_PROJECT_DIR` (or repo root).
  • Audit log JSON is OWASP A08-clean: no user-controlled keys; only

whitelisted `checkpoint` (number), `decision` (enum), `reason` (sanitized string).

  • Design doc patterns are read from `bkit.config.json` — no traversal

via untrusted `feature` arg (pattern is `{feature}` placeholder only).

  • Trust score read is fail-safe: missing file returns 0 (blocks engage).

Module Dependencies

| Module | Function | Usage | |--------|----------|-------| | `lib/control/fast-track.js` | `evaluatePreconditions(feature)` | Block list | | `lib/control/fast-track.js` | `engage(feature)` | Audit + record | | `lib/control/fast-track.js` | `recordCheckpoint({...})` | Per-CP decision | | `lib/control/fast-track.js` | `isCheckpointSkipped(n)` | Should we skip? |

Examples

# Engage on a Design-complete feature
/pdca-fast-track bkit-v2111-integrated-enhancement

# Result if Trust < 80
# Block: trust_score_too_low (current 68 < required 80)

# Result if Design missing
# Block: design_doc_missing → run /pdca design first

Related

  • `/control trust` — current Trust Score breakdown
  • `/control level <0-4>` — manual override
  • `/pdca-watch` — observe progress while fast-tracking
  • `/pdca design <feature>` — required prerequisite

ARGUMENTS:

Read more
Ships withbkit

A Claude Code plugin that verifies AI-generated code against its own design specs. Three commands. Anyone — even someone vibe-coding for the first time — can ship robust, production-quality software.

Get the whole plugin