Read Transcript

Search and retrieve full context from previous Angainor iteration transcripts.

When to Use

Use this skill when:

• •progress.txt mentions something but lacks detail
• •You need to understand HOW something was implemented, not just WHAT
• •You're debugging an issue that may relate to a previous iteration's work
• •You want to see the full conversation from a specific iteration

The Job

1. •Read the transcript index at transcripts/index.json
2. •Search for relevant transcripts by story ID, date, branch, or keyword
3. •Read the full transcript file(s) for detailed context
4. •Extract and present the relevant information

Index Format

The index at transcripts/index.json has this structure:

{
  "transcripts": [
    {
      "file": "2026-01-17-14-30-00-iteration-1.txt",
      "timestamp": "2026-01-17-14-30-00",
      "iteration": 1,
      "branch": "angainor/feature-name",
      "storyId": "US-001"
    }
  ]
}

Search Methods

By Story ID

Find all iterations that worked on a specific story:

Search transcripts for story US-003

By Date Range

Find iterations from a specific time period:

Search transcripts from 2026-01-15 to 2026-01-17

By Branch

Find all iterations for a feature branch:

Search transcripts for branch angainor/auth-system

By Iteration Number

Get a specific iteration:

Read transcript from iteration 5

Step-by-Step Process

1. •

   Read the index:

   code

   Read transcripts/index.json
   

2. •

   Find matching entries:

   • •Filter by the search criteria (storyId, timestamp, branch, iteration)
   • •List matching transcripts with their metadata
3. •

   Read relevant transcripts:

   code

   Read transcripts/[filename].txt
   

4. •

   Extract key information:

   • •What was implemented
   • •Key decisions made
   • •Errors encountered and how they were resolved
   • •Files that were modified

Example Usage

User asks: "What approach did the previous iteration use for the database migration?"

Your process:

1. •Read transcripts/index.json to find recent iterations
2. •Identify iterations that might involve database work (check storyId patterns)
3. •Read the relevant transcript file
4. •Extract and summarize the migration approach

Output Format

When presenting transcript information:

## Transcript: [filename]
**Iteration:** [N] | **Date:** [timestamp] | **Story:** [storyId]

### Summary
[Brief summary of what happened in this iteration]

### Key Details
- [Relevant detail 1]
- [Relevant detail 2]

### Files Modified
- [file1.ts]
- [file2.ts]

Tips

• •Start with the index to narrow down which transcripts to read
• •Transcripts can be long - focus on the sections relevant to the query
• •The metadata at the end of each transcript includes iteration number and branch
• •Cross-reference with progress.txt for additional context
• •Multiple iterations may have worked on the same story (if it failed and retried)

Checklist

Before responding:

• • Read transcripts/index.json first
• • Identified relevant transcripts based on search criteria
• • Read the full transcript file(s)
• • Extracted and presented the relevant information
• • Provided context about which iteration/story the information came from