Agent skill

continue-quickfix

Stars 15
Forks 1

Install this agent skill to your Project

npx add-skill https://github.com/leeovery/agentic-workflows/tree/main/skills/continue-quickfix

SKILL.md

Continue an in-progress quick-fix. Determines current phase and routes to the appropriate phase skill.

⚠️ ZERO OUTPUT RULE: Do not narrate your processing. Produce no output until a step or reference file explicitly specifies display content. No "proceeding with...", no discovery summaries, no routing decisions, no transition text. Your first output must be content explicitly called for by the instructions.

Instructions

Follow these steps EXACTLY as written. Do not skip steps or combine them.

CRITICAL: This guidance is mandatory.

  • After each user interaction, STOP and wait for their response before proceeding
  • Never assume or anticipate user choices
  • Complete each step fully before moving to the next

Step 0: Initialisation

Load casing-conventions.md and follow its instructions as written.

Run migrations — this is mandatory. You must complete it before proceeding.

Invoke the /workflow-migrate skill and follow its instructions exactly — if it issues a STOP gate, you must stop.

→ Proceed to Step 1.


Step 1: Discovery State

!node .claude/skills/continue-quickfix/scripts/discovery.cjs

If the above shows a script invocation rather than discovery output, the dynamic content preprocessor did not run. Execute the script before continuing:

bash
node .claude/skills/continue-quickfix/scripts/discovery.cjs

If discovery output is already displayed, it has been run on your behalf.

Parse the discovery output to understand:

From quick_fixes array:

  • name - the work unit name
  • next_phase - the phase to route to
  • phase_label - human-readable phase status
  • completed_phases - list of completed phases (for backwards navigation)

From top-level fields:

  • count - number of active quick-fixes
  • summary - human-readable state summary
  • completed / cancelled - arrays of non-active quick-fixes with name, status, last_phase
  • completed_count / cancelled_count - counts for each

IMPORTANT: Use ONLY this script for discovery. Do NOT run additional bash commands (ls, head, cat, etc.) to gather state.

→ Proceed to Step 2.


Step 2: Check Count and Arguments

If count is 0

Output the next fenced block as a code block:

Continue Quick-Fix

No quick-fixes in progress.

Run /start-quickfix to begin a new one.

STOP. Do not proceed — terminal condition.

If work_unit argument $0 provided

Store the work_unit.

→ Proceed to Step 4.

If work_unit not provided

→ Proceed to Step 3.


Step 3: Select Quick-Fix

Load select-quickfix.md and follow its instructions as written.

→ Proceed to Step 4.


Step 4: Validate Selection

Load validate-selection.md and follow its instructions as written.

→ Proceed to Step 5.


Step 5: Backwards Navigation

Load revisit-phase.md and follow its instructions as written.

→ Proceed to Step 6.


Step 6: Route to Phase Skill

Using the selected quick-fix's next_phase, invoke the appropriate phase skill:

next_phase Invoke
scoping /workflow-scoping-entry quick-fix {work_unit}
implementation /workflow-implementation-entry quick-fix {work_unit}
review /workflow-review-entry quick-fix {work_unit}

Skills receive positional arguments: $0 = work_type (quick-fix), $1 = work_unit. Topic is inferred from work_unit.

If the user chose to revisit a completed phase in Step 5, use that phase instead of next_phase.

Invoke the skill. This is terminal — do not return to the backbone.

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results