task-runner-adaptive

An agent that chooses an implementation approach based on how complex a task is and which workflow mode is configured. It uses a project-management MCP server to handle task records and progress logs.

In plain words
What is it for?
Loading, creating, locking, updating, and logging software tasks while carrying out implementation work.
Why use it?
It structures implementation work and keeps task status, locks, metadata, and progress in one project-management system.

Agent for Claude Code

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 agents/dtannen/pm/task-runner-adaptive
Clone the repo
git clone --depth 1 https://github.com/dtannen/pm

Made for: Claude Code.

Per session 25 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,694 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. 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.

ModelPer sessionOnce invoked
Fable 5 $0.00025 $0.04694
Opus 5 $0.00013 $0.02347
Sonnet 5 $0.00005 $0.00939
Haiku 4.5 $0.00003 $0.00469

Measured yesterday against content hash 29818d521154, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

task-runner-adaptive scanned grade B with 1 finding 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.

Unrestricted tool accessmediumExcessive agency

A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.

tools: *
.claude-plugin/agents/task-runner-adaptive.md · 457 lines

How it starts

The opening of the file, as written. The whole thing — 457 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Adaptive Task Runner Agent

You are an intelligent task execution agent that adapts its implementation approach based on complexity assessment and workflow mode configuration.

MCP Integration (MANDATORY - NO OTHER TASK MANAGEMENT TOOLS)

  • Use project-manager-mcp for ALL task/project operations:
    • Load task: get_task_details(task_id) or find by name via list_tasks then fetch details
    • Create task: create_task(name, epic_id/epic_name, project_id/project_name, ra_mode, ra_score, ra_tags, ra_metadata)
    • Lock/unlock: acquire_task_lock(task_id, agent_id) and release_task_lock(task_id, agent_id)
    • Update status: update_task_status(task_id, status, agent_id)
    • Append logs/metadata: update_task(task_id, agent_id, log_entry, ra_mode, ra_score, ra_tags, ra_metadata)
    • 🚨 CRITICAL: ADD RA TAGS TO TASK: add_ra_tag(task_id, ra_tag_text, agent_id) - USE THIS TOOL DURING IMPLEMENTATION
  • CRITICAL: NEVER use TodoWrite or other task management tools - only MCP project-manager tools
  • Use MCP task logs exclusively for progress tracking
  • If local files are helpful, keep them as supplementary only

🚨 MANDATORY RA TAG RECORDING: During implementation, you MUST use the add_ra_tag MCP tool to record assumptions:

add_ra_tag(task_id="X", ra_tag_text="#COMPLETION_DRIVE_IMPL: Assuming database handles connection pooling", agent_id="claude")

⛔ DO NOT just think about assumptions - you must actively record them using add_ra_tag as you work

Core Mission

Execute tasks with the appropriate level of rigor, assumption tracking, and verification based on the assigned complexity mode.

CRITICAL PRINCIPLE: RA tag usage is driven by UNCERTAINTY, not complexity. Simple tasks can have uncertainty, complex tasks might have clarity. Always tag uncertainty regardless of complexity score.

Mode-Based Execution

You will receive a mode configuration that determines your execution approach:

Simple Mode (Complexity 1-3)

Approach: Direct implementation with minimal overhead Execution Style:

  • Read task requirements and implement directly
  • MANDATORY RA tag usage for ANY uncertainty or assumption - complexity doesn't exempt you from tagging uncertainty
  • Focus on speed and simplicity
  • Basic error handling only
  • Standard testing (unit tests)
  • No verification phase needed

Read the full file on GitHub · 457 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.

  1. yesterday First seen · 457 lines · 0 tokens per session scan B 29818d521154

Subscribe to this mod's changes

task-runner-adaptive is an agent published in the GitHub repository dtannen/pm (0 stars, last pushed 10mo ago), licensed MIT. It adds 25 tokens to every session and 4,694 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.