agentic-tpm

Solid

Coordinate cross-functional project delivery by tracking commitments, dependencies, decisions, and risks; maintain project context and prepare bounded follow-ups and status reports. Use for an ongoing TPM or project-coordination role, not implementation or personnel management.

AI & Automation 156 stars 14 forks Updated yesterday MIT

Install

View on GitHub

Quality Score: 83/100

Stars 20%
73
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# Agentic TPM Own the coordination loop: ask, observe, document, and report. Help the human sponsor drive delivery without claiming authority over people, budgets, scope, or release decisions. Credit: inspired by Steve Yegge's Agentic TPM proposal supplied by the user; see [research and attribution](references/research.md). ## Start with a project lane Use existing context first. Establish the outcome and acceptance criteria, human sponsor/decision-maker, scope and exclusions, target dates, known owners, permitted sources, and a private local project folder. Ask only for missing information that changes the next action. A vague project is enough to start a local draft with explicit unknowns; do not fabricate commitments. Use [the project record](references/project-record.md) to initialize one `projects/<project-id>/PROJECT.md` and `OUTBOX.md`, or adapt the user's existing records instead of creating a competing tracker. Never overwrite existing project records. Read the charter and outstanding items when resuming; keep project lanes separate. Use stable IDs and source links so another TPM can understand and reconcile the records. Default to **draft-only** external communication. Record a communication mandate only when the operator explicitly grants it: allowed recipients, channels, message purposes, permitted information, cadence, quiet hours/time zone, escalation route, expiry, and revocation. Existing specific authorization is sufficient within its scope; do not ask a...

Details

Author
HybridAIOne
Repository
HybridAIOne/hybridclaw
Created
7 months ago
Last Updated
yesterday
Language
TypeScript
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category

AI & Automation Listed

ppm-agent

The strategic brain of the PMO — reads any project artifact and pushes every actionable item toward resolution. Use when uploading transcripts, asking about project status, needing decisions framed, or requesting risk assessment. Triggers: "review this", "what's the state of [project]", "process this transcript", "triage this", "what needs my attention", "what actions came out of this", "what needs to surface."

0 Updated today
cody-hutson
AI & Automation Listed

tpm-task

Use to capture, track, and manage a project's lightweight task ledger — a shared, hand-editable markdown to-do list Claude and the user both read. `/tpm-task <mode> <freeform>` where mode is add · import · export · list · show · edit · check · add-subtask · start · finish · drop · remove · reopen. Bare "/tpm-task" or "list" shows the open tasks. The tool owns all mechanics; this skill owns judgment (draft content, detect epics, confirm destructive ops). NOT a real issue tracker — keep it skimmable.

0 Updated today
codercowboy
AI & Automation Listed

a2a-manager-agent-orchestration

Use this skill to operate a long-running project through multiple persistent agent sessions. Trigger when a user appoints a manager or director, coordinates named worker sessions, uses session-to-session A2A communication, shares one repository or workspace, manages context compaction or session handoff, or needs architecture and rules to survive across many tasks. Preserve global intent in a manager session, stable domain context in worker sessions, and durable project state outside every chat. Do not trigger for ordinary one-shot tasks, ephemeral subagent calls, or merely because agent tools are available.

0 Updated 3 weeks ago
EESIZ