feat(interrogate): add interrogate adversarial-review skill plugin

.factory-plugin/marketplace.json

@@ -53,6 +53,12 @@
       "description": "Pull request lifecycle skills: create PRs with consistent conventions and follow up on them until merge-ready",
       "source": "./plugins/code-review",
       "category": "productivity"
+    },
+    {
+      "name": "interrogate",
+      "description": "Adversarial review where only findings proven by a run against the real code survive",
+      "source": "./plugins/interrogate",
+      "category": "quality"
     }
   ]
 }

README.md

@@ -96,6 +96,14 @@ Skills for continuous learning and improvement.
 
 Autonomous experiment loop for optimization research. Try an idea, measure it, keep what works, discard what doesn't, repeat. Works standalone or as a mission worker.
 
+### interrogate
+
+Adversarial review where a finding survives only when a run proves it.
+
+**Skills:**
+
+- `interrogate` - Fan out adversaries to break a change or design; report only verified findings
+
 ## Plugin Structure
 
 Each plugin follows the Factory plugin format:

plugins/interrogate/.factory-plugin/plugin.json

@@ -0,0 +1,8 @@
+{
+  "name": "interrogate",
+  "description": "Adversarial review where only findings proven by a run against the real code survive",
+  "author": {
+    "name": "yifu zuo",
+    "email": "yzuo.2019@gmail.com"
+  }
+}

plugins/interrogate/README.md

@@ -0,0 +1,9 @@
+# interrogate
+
+Adversarial review where a finding survives only when a run proves it.
+
+**Skills:**
+
+- `interrogate` - Fan out adversaries to break a change or design; report only findings verified by rerunning their commands
+
+Adapted from the pstack skills by Lauren Tan ([cursor/plugins](https://github.com/cursor/plugins), MIT License), with the fan-out rewritten to be harness-agnostic.

plugins/interrogate/skills/interrogate/SKILL.md

@@ -0,0 +1,30 @@
+---
+name: interrogate
+description: "Use for \"interrogate\", \"adversarial review\", \"multi-model review\", \"challenge this\", \"stress test this code\", \"find blind spots\", or \"tear this apart\". Adversaries try to break a change or design, and only findings proven by a run survive."
+---
+
+# Interrogate
+
+Adversaries try to break a change or a design. A finding survives only when a run proves it. Report the findings. Do not apply fixes.
+
+## Frame
+
+The scope is what the user names. Otherwise it is the branch diff from its merge base with the base branch, through the working tree. State the intent in your own words, drawn from the user's message, the commits, and the PR description. Ask the user when the intent is unclear.
+
+## Fan out
+
+Spawn several adversary subagents in isolated worktrees, ideally on different model families; copy the scope's uncommitted changes into each worktree. Give each adversary the scope, the intent, and this brief:
+
+> Break this change. Prove each finding with a run against the real code, such as a failing test, a repro script, or command output that shows the wrong behavior. For a design or plan, run a prototype or probe the current tree. Keep repros in your own worktree. Report each finding with a title, the exact command, the output it printed, and the output the intent requires. Leave out anything you did not run. That covers style, naming, structure, and risks that need a future change to matter. No findings is a valid result.
+
+When subagents or a second model are unavailable, run the review serially in a fresh context and say so.
+
+## Verify
+
+Rerun every finding's command yourself. Keep a finding when the output shows the failure and the failure breaks the intent. Drop the rest. Merge findings that share a root cause.
+
+## Output
+
+- **Intent**: the restatement.
+- **Findings**: for each verified finding, the command, its output, how it breaks the intent, and the adversaries that found it. Say "no verified findings" when none survive.
+- **Dropped**: for each adversary, the count of dropped findings and why.

skills/interrogate

@@ -0,0 +1 @@
+../plugins/interrogate/skills/interrogate