Today's Agent Skill: Turn Any Message Thread Into a Decision Log

What It Does

Long Slack threads and email chains bury decisions inside chatter. Two weeks later nobody remembers what was actually agreed, who owns it, or whether it shipped. This skill mines a thread and emits a clean decision log.

How It Works

The agent ingests raw thread text, separates decisions from discussion, and attributes each decision to the person who made it with a timestamp anchor. Open questions and implicit commitments get pulled into a separate section so nothing hides in the noise.

How to Deploy It

Save the SKILL.md content below to your agent's skills directory as `inbox-to-action-log/SKILL.md` and reload the agent. Paste any thread and say "log the decisions" — the trigger phrases fire automatically.

SKILL.md — Ready to Deploy

## Description
Extracts decisions, owners, and open questions from any multi-participant message thread (email chain, Slack channel, ticket comments, meeting chat) and emits a structured decision log. Distinguishes committed decisions from speculation, and flags commitments made without an explicit owner.

## Trigger
Fires on: log the decisions, what was decided, decision log, extract decisions, who agreed to what, pull the commitments, what did we actually decide. Also fires when the user pastes a thread longer than roughly 10 messages and asks any question about outcomes, even without naming the skill.

## Input
- Raw thread text (any format — pasted email chain, exported Slack, ticket comments)
- Optional: a date range to bound the extraction
- Optional: a list of participant names to normalize handles against

If the thread lacks timestamps, proceed and note the gap in Output. Do not ask the user to reformat.

## Steps
1. Parse the thread into discrete messages with author and timestamp where available.
2. Classify each message as: decision, proposal, question, status update, or noise. Discard noise.
3. For each decision, identify the deciding author and the last message that c

Copy the full SKILL.md and drop it into your agent's skills directory to activate this skill.

← Back to All Posts | Home — MD AI Tools