A task-start command that assesses how difficult a project task is and chooses a suitable way to carry it out. It can look up the task, check dependencies, lock it for one worker, update its status, and record progress.

Use it to begin named tasks in a project-management system while checking prerequisites and keeping their progress updated.

It removes the need to decide the workflow and manage task state by hand before starting work. Locking helps prevent two workers from taking the same task.

task-start — Command by dtannen

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/dtannen/pm/task-start

Clone the repo

git clone --depth 1 https://github.com/dtannen/pm

Made for: Claude Code.

Per session 12 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 2,713 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	$0.00012	$0.02713
Opus 5	$0.00006	$0.01357
Sonnet 5	$0.00002	$0.00543
Haiku 4.5	$0.00001	$0.00271

Measured yesterday against content hash 270ccdcde62c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

task-start 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-plugin/commands/task-start.md · 374 lines

How it starts

The opening of the file, as written. The whole thing — 374 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Adaptive Task Start

Intelligent task execution with automatic complexity assessment and adaptive workflow selection.

Usage

/pm:task-start <task-name>

Example: /pm:start user-authentication

Instructions

You are launching the PM Adaptive workflow - an intelligent system that automatically assesses task complexity and selects the optimal execution mode.

MCP Integration (preferred)

Use MCP for task lookup, locking, status changes, and progress logging.

Full name: mcp__project-manager-mcp__list_tasks
Full name: mcp__project-manager-mcp__get_task_details
Full name: mcp__project-manager-mcp__acquire_task_lock
Full name: mcp__project-manager-mcp__update_task_status
Full name: mcp__project-manager-mcp__update_task
Full name: mcp__project-manager-mcp__release_task_lock
Full name: mcp__project-manager-mcp__create_task (if creating tasks on the fly)

MCP flow:

Resolve the task by name: mcp__project-manager-mcp__list_tasks with a filter, then ...__get_task_details.
Validate dependencies from get_task_details (use ra_metadata/dependencies fields if present).
Acquire a lock: mcp__project-manager-mcp__acquire_task_lock(task_id, agent_id).
Mark IN_PROGRESS: mcp__project-manager-mcp__update_task_status(task_id, IN_PROGRESS, agent_id).
Persist assessment results and selected mode: mcp__project-manager-mcp__update_task(ra_score, ra_mode, log_entry, ra_metadata).
Log execution milestones via repeated ...__update_task(log_entry=...) calls.
On completion, move to REVIEW or DONE using ...__update_task_status and release the lock with ...__release_task_lock.

1. Task Validation

Prefer MCP: lookup task by name using mcp__project-manager-mcp__list_tasks and confirm via ...__get_task_details.
Fallback: check .pm/tasks/$ARGUMENTS.md exists.
If task is already completed, prompt for override.

2. Dependency Check

Verify dependencies:

Prefer MCP: read dependencies or ra_metadata.depends_on via ...__get_task_details and ensure each dependent task is completed.
Fallback: parse .pm/tasks/$ARGUMENTS.md frontmatter.
If unmet: block start and list missing dependencies.

Read the full file on GitHub · 374 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 · 374 lines · 12 tokens per session scan A 270ccdcde62c

Subscribe to this mod's changes

task-start is a command published in the GitHub repository dtannen/pm (0 stars, last pushed 10mo ago), licensed MIT. It adds 12 tokens to every session and 2,713 once invoked, about $0.0001 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.

Planning Productivity

task-start

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Adaptive Task Start

Usage

Instructions

MCP Integration (preferred)

1. Task Validation

2. Dependency Check

What this file has done since we first saw it

Other commands, from other repositories

checklist

clarify

specify

analyze

converge

implement