parallel-data-enrichment

Files (1)

• SKILL.md

SKILL.md (first lines)

> **Note:** This research was generated using AI assistance (Claude + Parallel.ai) with human expert review. See [methodology](../../docs/methodology.md) for details.

## Data Enrichment

Enrich: $ARGUMENTS

### Step 1: Start enrichment

```bash
parallel-cli enrich run "$ARGUMENTS" --no-wait --json
```

For follow-ups chaining a prior interaction:

```bash
parallel-cli enrich run "$ARGUMENTS" --no-wait --json --previous-interaction-id "$INTERACTION_ID"
```

Parse the JSON output to extract `taskgroup_id`, `interaction_id`, and the monitoring URL. Tell the user enrichment has started and share the monitoring URL.

### Step 2: Poll for results

```bash
parallel-cli enrich poll "$TASKGROUP_ID" -o "$OUTPUT_FILE.csv" --timeout 540
```

- Use `--timeout 540` to stay within tool execution limits
- Results are written to the specified CSV output file
- Share the `interaction_id` for follow-up enrichment chaining

### Response format

After completion:
1. Confirm the output CSV path
2. Share the `interaction_id` for follow-up questions
3. Summarize what fields were enriched

### Setup

```bash
curl -fsSL https://parallel.ai/install.sh | bash
# or
pipx install "parallel-web-tools[cli]" && pipx ensurepath
parallel-cli login
# or: export PARALLEL_API_KEY="your-key"
```

From hack4rva/pillar-thriving-city-hall, Apache-2.0.