slack-context-fetcher

Files (1)

• SKILL.md

SKILL.md (first lines)

# Slack Context Fetcher

Fetches Slack thread content when tickets reference Slack discussions. Uses the Pipeline API for state management.

## Configuration

Uses the **production API** by default. No configuration needed for read operations.

**Defaults (no setup required):**

- API URL: `https://api-gateway-856475788601.us-central1.run.app`
- Read-only endpoints at `/public/*` require no authentication

**For write operations** (state transitions), set:

```bash
export PIPELINE_API_KEY="..."  # Get from GCP Secret Manager or ask admin
```

## Prerequisites

- `slackdump` CLI installed at `/usr/bin/slackdump`
- Credentials stored at `~/.cache/slackdump/comfy-organization.bin`
- If auth fails, load `slackdump` skill and read `reference/auth.md`

## Workflow

### 1. Detect Slack References

Parse ticket description for Slack links:

```
Pattern: https://{workspace}.slack.com/archives/{channel_id}/p{timestamp}
Example: https://comfy-org.slack.com/archives/C07ABCD1234/p1234567890123456
```

Extract:

- `channel_id`: C07ABCD1234
- `thread_ts`: 1234567890.123456 (insert decimal before last 6 digits)

### 2. Fetch Thread Content via slackdump

Use the `slackdump` skill's **export-thread** workflow:

```bash
THREAD_URL="https://comfy-organization.slack.com/archives/$CHANNEL/$THREAD_TS_NO_DOT"

# Export the thread
slackdump dump "$THREAD_URL"

# Or using colon notation
slackdump dump "$CHANNEL:$THREAD_TS"
```

> **Note:** Load the `slackdump` skill for detailed instructions. Use the `reference/export-thread.md` workflow.

### 3. Resolve User Names

User names are included in slackdump export output. If additional resolution is needed:

```bash
# List users in the workspace
slackdump list users
```

> Load the `slackdump` skill and use `reference/discover.md` for user lookup.

### 4. Format for Context

```markdown
## Slack Thread Context

**Thread:** [Link](https://comfy-org.slack.com/archives/C07.../p1234...)
**Participants:** Alice, Bob, Charlie
**Messages:** 12

### Full Thread

**Alice** (2024-01-15 10:30):

> Original message about the issue...

**Bob** (2024-01-15 10:45):

> Response with context...
```

### 5. Save Artifacts and Update State

Save the Slack context to the local runs directory and update ticket metadata via API:

```bash
# Save local artifact
echo "$SLACK_CONTEXT" > "$RUN_DIR/slack-context.md"
```

```typescript
import { PipelineClient } from '@pipeline/client'

const client = new PipelineClient({
  apiUrl: process.env.PIPELINE_API_URL || 'https://api-gateway-856475788601.us-central1.run.app',
  agentId: process.env.AGENT_ID,
})

// Verify ticket exists - fail loudly if ticket-intake wasn't run
const ticket = await client.getTicket(ticketId).catch(() => null)
if (!ticket) {
  console.error('❌ No pipeline ticket found. Run ticket-intake first to register this ticket.')
  // Stop here - ticket-intake must run before this skill
}
// Update ticket metadata with Slack thread info
await client.updateTicket(ticketId, {
  metadata: {
    ...ticket.metadata,
    slackThreads: [
      ...(ticket.metadata?.slackThreads || []),
      { url: threadUrl, fetched: new Date().toISOString(), messages: messageCount },
    ],

From DrJKL/ticket-to-pr-pipeline, MIT.

More from DrJKL/ticket-to-pr-pipeline