A task-planning command that turns a development request into confirmed project notes and a step-by-step plan. It researches the codebase and saves task, architecture, affected-area, and plan documents.

Use it to start a coding task, especially in an unfamiliar repository. It initializes the planning workspace, maps the codebase, checks the technology stack, and records the planned work.

It gives later implementation steps a shared understanding of the requirements, project structure, and likely files to change.

fd-task — Command by DVNghiem

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/dvnghiem/flowdeck/fd-task

Clone the repo

git clone --depth 1 https://github.com/DVNghiem/FlowDeck

Per session 33 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 2,455 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	$0.00033	$0.02455
Opus 5	$0.00016	$0.01228
Sonnet 5	$0.00007	$0.00491
Haiku 4.5	$0.00003	$0.00246

Measured yesterday against content hash d7ab8c728172, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

fd-task 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.

docs/commands/fd-task.md · 326 lines

How it starts

The opening of the file, as written. The whole thing — 326 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Task

Pipeline entrypoint. Turns a task description into the four confirmed artifacts every later stage reads.

Input: $ARGUMENTS — the task description. REQUIRED.

If $ARGUMENTS is empty, the agent MUST prompt the user for a task description before doing anything else. The agent MUST NOT guess or infer the task.

Step 1: Auto-init

Check whether ~/.fd-plan/<slug>/ exists, where <slug> is the project directory name.

MUST initialize if missing:

Create ~/.fd-plan/<slug>/.
Map the codebase with the graph:
- Run fdx-graph action:status — it reports build age without paying for a build.
- Absent or stale → run fdx-graph action:build. A no-op build is cheap and leaves the cache untouched.
- Run fdx-graph action:report and read the generated GRAPH_REPORT.md for god nodes and cluster orientation.
- For anything the graph does not cover (tech stack, dependency versions), delegate to @mapper or read package.json / go.mod / Cargo.toml / pyproject.toml plus the src/ tree.
Write ~/.fd-plan/<slug>/architecture.md — the project-level tech design: tech stack, module layout, entry points, established conventions, external dependencies.
Initialize STATE.md via planning-state action:update with createDefaultState() values, and create ~/.fd-plan/<slug>/config.json with the default config.

Log: "Initialized ~/.fd-plan/<slug>/ — project architecture mapped."

If it already exists, skip init. The agent MUST NOT overwrite an existing architecture.md.

Step 2: Research gate

Before searching anything, analyze the task description and propose what to research.

2a. Propose queries

From the task description, derive 3-5 specific research queries. Each query should target a distinct area (e.g. existing implementation, relevant dependencies, affected modules, prior decisions, external docs).

The agent MUST present them to the user:

Research plan for: "<task description>"

Proposed queries:
  1. <query>
  2. <query>
  3. <query>

[Y] Run these queries
[N] Skip research — go straight to discussion
[C] Use custom queries

Read the full file on GitHub · 326 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 · 326 lines · 33 tokens per session scan A d7ab8c728172

Subscribe to this mod's changes

fd-task is a command published in the GitHub repository DVNghiem/FlowDeck (24 stars, last pushed 13d ago), licensed MIT. It adds 33 tokens to every session and 2,455 once invoked, about $0.0002 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.

Orchestration Planning Research Writing and docs agentic-ai ai-agents opencode opencode-plugin

fd-task

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Task

Step 1: Auto-init

Step 2: Research gate

What this file has done since we first saw it

Other commands, from other repositories

annotate-plan

wtfp:create-outline

wtfp:execute-outline

wtfp:pause-writing

wtfp:resume-writing

wtfp:review-section