ix-impact

> [ix-claude-plugin shared model](../shared.md) Check `command -v ix` first. If unavailable, use Grep to find all usages and estimate impact manually. ## Argument parsing Strip `--save` and any following path token from `$ARGUMENTS` before resolving the target. - If `--save <path>` is present, set `SAVE_PATH` to that path. - If `--save` is present without a path, auto-generate `ix-impact-<target-slug>.md` in cwd (target slug = the target with spaces and slashes replaced by `-`). - If `--save` is absent, `SAVE_PATH` is empty — do not write a file. ## Pro check (optional) Do not run `ix briefing` — you have already been told. The plugin's UserPromptSubmit hook injects an `[ix] Session briefing` once per TTL whenever Pro is reachable. Seen one this session? Pro is available; read `openBugs` from it. Not seen one? Skip every **[Pro]** step. Asking costs a whole turn to learn something already in your context. ## Goal Answer: *what breaks if this changes, and is it safe to proceed?* Stop as early as the risk level allows. ## Phase 1 — Risk score (always) Run in parallel: ```bash ix impact $ARGUMENTS --format llm ix explain $ARGUMENTS --format llm ``` **God-module check:** If `fan_out > 20 AND fan_in < 2` in the `ix explain` result: > ⚠ This symbol has high fan_out and low fan_in — it reaches out to many dependents but has few callers. Standard blast-radius metrics may understate risk. Check callers of its key dependencies, not just direct dependents. This caveat applies regardless of the `ix impact` risk classification. **Immediately classify:** | Risk level | Action | |---|---| | `low` + < 3 dependents | **STOP** — safe to proceed. Report and suggest verification targets. | | `medium` OR 3–10 dependents | Go to Phase 2 | | `high` or `critical` OR > 10 dependents | Go to Phase 2 + 3 | ## Phase 2 — Callers and dependents (medium/high/critical) Run in parallel: ```bash ix callers $ARGUMENTS --limit 15 --format llm ix depends $ARGUMENTS --depth 2 --format llm ``` Extract: direct callers by name and subsystem, transitive count. **Stop here if risk is `medium`:** report callers, suggest verification, done. ## Phase 3 — Import chain and subsystem spread (high/critical only) ```bash ix imported-by $ARGUMENTS --format llm ``` Cross-reference callers + dependents + importers to identify: - Which subsystems are in the blast radius - Whether the change crosses an architectural boundary - Any tests that cover the affected paths ## Output ``` ## Impact: [target] **Risk level:** <critical | high | medium | low> **Verdict:** <SAFE TO PROCEED | REVIEW CALLERS FIRST | NEEDS CHANGE PLAN> **Blast radius:** - Direct dependents: N - Transitive (depth 2): M - Subsystems affected: [list — only if phase 3 ran] **Key callers:** [top 5, with subsystem label] **At-risk behaviors:** [from the ix impact behavior lines] **Recommended action:** - low: proceed, verify [specific callers] - medium: test [caller list] after change - high/critical: run `/ix-plan $ARGUMENTS` before editing **Known bugs in blast radius:** [Pro only — list open bugs touching callers/dependents, or omit section if none/Pro unavailable] ``` ## Phase 4 — Known bugs in blast radius **[Pro]** If Pro is available and `openBugs` from the briefing is non-empty, check for bugs affecting this target: ```bash ix bug list --format text ``` Cross-reference open bugs against the direct callers and dependents identified in Phase 2. Any open bug touching the blast radius escalates the risk verdict — flag it explicitly in the output. Never read source code in this skill. Risk analysis is purely graph-based. **Save step (only if `SAVE_PATH` is non-empty):** - Write the full output above to `SAVE_PATH` using the Write tool. - Confirm to the user: `Saved to <SAVE_PATH>`. - Do not write the file if `--save` was not passed.