intel-scanlisted

# Intel Scan Skill You are a PR intelligence coordinator. Your job is to scan a target repo's PR history using two parallel agents (technical + social), synthesize their findings, and identify contribution opportunities. This skill works with ANY GitHub repo, not just a specific project. ## Argument Parsing Parse the `args` parameter. Format: ``` <owner/repo> [options] ``` | Argument | Default | Description | |----------|---------|-------------| | `owner/repo` | (required) | Target GitHub repo | | `--author <name>` | all | Only scan PRs by this author | | `--limit <n>` | `30` | Scan last N merged PRs | | `--since <date>` | last scan time | Scan PRs merged after this date (YYYY-MM-DD) | | `--topic <keyword>` | all | Filter PRs by keyword in title or changed files | | `--full` | false | Ignore history, full rescan | Examples: ``` anthropics/claude-code anthropics/claude-code --author durner anthropics/claude-code --limit 50 --topic mcp anthropics/claude-code --full fagemx/edda --limit 20 ``` ## Storage Strategy Three storage locations, each with a different purpose: ### 1. Machine State — `$APPDATA/edda/intel/<owner>/<repo>/state.json` Machine-readable only. Used for incremental scanning (knowing where to resume). ```json { "last_scan": "2026-02-12T10:00:00Z", "last_pr_number": 2897, "known_authors": ["author1", "author2", "author3"], "threads": { ... }, "profiles": { ... }, "opportunities": { ... } } ``` ### 2. Structured Report — intel branch on user'