A command that turns an approved feature specification into a list of implementation tasks. It reads the requirements, technical design, and project guidance before creating the task document.

Use it to generate or update a tasks.md file for a Kiro feature specification, with an option to approve the requirements and design automatically.

It breaks a feature into concrete coding work and checks that the necessary planning documents are present first.

spec-tasks — Command by kbwo

Install

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.

agentmods

npx agentmods add commands/kbwo/ccmanager/spec-tasks

Clone the repo

git clone --depth 1 https://github.com/kbwo/ccmanager

Made for: Claude Code.

Per session 6 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,747 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What 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	$0.00006	$0.01747
Opus 5	$0.00003	$0.00873
Sonnet 5	$0.00001	$0.00349
Haiku 4.5	$0.00001	$0.00175

Measured yesterday against content hash 59215aa9b6d7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

spec-tasks 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 yesterday.

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.

.claude/commands/kiro/spec-tasks.md · 177 lines

How it starts

The opening of the file, as written. The whole thing — 177 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Implementation Tasks

Generate detailed implementation tasks for feature: $1

Task: Generate Implementation Tasks

Prerequisites & Context Loading

If invoked with -y flag ($2 == "-y"): Auto-approve requirements and design in spec.json
Otherwise: Stop if requirements/design missing or unapproved with message: "Run /kiro:spec-requirements and /kiro:spec-design first, or use -y flag to auto-approve"
If tasks.md exists: Prompt [o]verwrite/[m]erge/[c]ancel

Context Loading (Full Paths):

.kiro/specs/$1/requirements.md - Feature requirements (EARS format)
.kiro/specs/$1/design.md - Technical design document
.kiro/steering/ - Project-wide guidelines and constraints:
- Core files (always load):
  - @.kiro/steering/product.md - Business context, product vision, user needs
  - @.kiro/steering/tech.md - Technology stack, frameworks, libraries
  - @.kiro/steering/structure.md - File organization, naming conventions, code patterns
- Custom steering files (load all EXCEPT "Manual" mode in AGENTS.md):
  - Any additional *.md files in .kiro/steering/ directory
  - Examples: api.md, testing.md, security.md, etc.
- (Task planning benefits from comprehensive context)
.kiro/specs/$1/tasks.md - Existing tasks (only if merge mode)

CRITICAL Task Numbering Rules (MUST FOLLOW)

⚠️ MANDATORY: Sequential major task numbering & hierarchy limits

Major tasks: 1, 2, 3, 4, 5... (MUST increment sequentially)
Sub-tasks: 1.1, 1.2, 2.1, 2.2... (reset per major task)
Maximum 2 levels of hierarchy (no 1.1.1 or deeper)
Format exactly as:

- [ ] 1. Major task description
- [ ] 1.1 Sub-task description
  - Detail item 1
  - Detail item 2
  - _Requirements: X.X, Y.Y_

- [ ] 1.2 Sub-task description
  - Detail items...
  - _Requirements: X.X_

- [ ] 2. Next major task (NOT 1 again!)
- [ ] 2.1 Sub-task...

Task Generation Rules

Natural language descriptions: Focus on capabilities and outcomes, not code structure
- Describe what functionality to achieve, not file locations or code organization
- Specify business logic and behavior, not method signatures or type definitions
- Reference features and capabilities, not class names or API contracts
- Use domain language, not programming constructs
- Avoid: File paths, function/method names, type signatures, class/interface names, specific data structures
- Include: User-facing functionality, business rules, system behaviors, data relationships
- Implementation details (files, methods, types) come from design.md
Task integration & progression:
- Each task must build on previous outputs (no orphaned code)
- End with integration tasks to wire everything together
- No hanging features - every component must connect to the system
- Incremental complexity - no big jumps between tasks
- Validate core functionality early in the sequence
Flexible task sizing:
- Major tasks: As many sub-tasks as logically needed
- Sub-tasks: 1-3 hours each, 3-10 details per sub
- Group by cohesion, not arbitrary numbers
- Balance between too granular and too broad
Requirements mapping: End details with _Requirements: X.X, Y.Y_ or _Requirements: [description]_
Code-only focus: Include ONLY coding/testing tasks, exclude deployment/docs/user testing

Read the full file on GitHub · 177 lines

Changes

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.

yesterday First seen · 177 lines · 6 tokens per session scan A 59215aa9b6d7

Subscribe to this mod's changes

spec-tasks is a command published in the GitHub repository kbwo/ccmanager (1,229 stars, last pushed yesterday), licensed MIT. It adds 6 tokens to every session and 1,747 once invoked, about $0.0000 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 2026-08-30.

Documents Planning claude claude-code cli gemini gemini-cli vibe-coding

spec-tasks

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Implementation Tasks

Task: Generate Implementation Tasks

Prerequisites & Context Loading

CRITICAL Task Numbering Rules (MUST FOLLOW)

Task Generation Rules

What this file has done since we first saw it

Other commands, from other repositories

package-assistant

implement

adversarial-review

ask

rescue

review