feat: restructure into a multi-extension plugin framework for Claude Code and agy

Replace the skills-only SSoT (skills/<name>/SKILL.md) with plugins/<name>/,
where a plugin can bundle skills, MCP servers, lifecycle hooks, rules,
commands, and agents. scripts/generate_plugins.py now discovers plugins
from one or more layered --source roots and emits real per-target trees
(dist/claude-code/** + .claude-plugin/marketplace.json, dist/agy/**),
translating hooks.json/mcp.json between Claude Code's and agy's actual
schemas instead of only fanning out flat skill lists.

Adds a private-overlay build path (--source/--private-repo/--out/
--install-local) so a second, non-public repo (PII, tokens, personal-only
plugins) can extend or override the public plugin set without either repo
touching the other's history — private builds default to a gitignored
dist-private/ and never land in tracked output.

Drops install.sh and descriptions.json: both only ever covered skills and
pre-date the native plugin marketplace; both tools now install the plugin
marketplace the documented way (Claude Code's /plugin marketplace add,
agy's plugins.json entries). CI path filters move to plugins/**, and the
workflow gains workflow_dispatch for a manual re-trigger.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014YVWWAnvR5TDQzmfY3RXG9

plugins/systematic-enumeration/plugin.json

@@ -0,0 +1,6 @@
+{
+  "name": "systematic-enumeration",
+  "description": "Forces element-by-element verification for finite sets to prevent counting errors.",
+  "version": "1.0.0",
+  "author": "Rootiest"
+}

plugins/systematic-enumeration/skills/systematic-enumeration/SKILL.md

@@ -0,0 +1,40 @@
+---
+name: systematic-enumeration
+description: Forces element-by-element verification for finite sets to prevent counting errors.
+version: 1.0.0
+user-invocable: true
+author: Rootiest
+---
+
+# Systematic Enumeration & Verification Skill
+
+## **Objective**
+To eliminate heuristic errors and "hallucinated patterns" when analyzing finite sets. This protocol overrides the model's tendency toward "holistic recognition" in favor of systematic, element-by-element verification.
+
+## **Execution Protocol**
+When this skill is triggered, you MUST NOT provide a direct answer immediately. Follow these three phases to ensure accuracy:
+
+### **Phase 1: Set Definition**
+Explicitly define the boundaries and members of the finite set being analyzed. 
+*   **Requirement:** List the members before performing any tests.
+*   *Example:* "The set consists of the files in the `/src` directory: [main.rs, utils.rs, types.rs]."
+
+### **Phase 2: Atomic Element Testing (O(n))**
+Iterate through every item in the set. For each item, perform a literal check against the target property.
+*   **Format:** Use a list or table to force token-level focus on each element.
+*   **Structure:** `[Item] -> [Logic/Observation] -> [Boolean Result]`
+*   *Note:* For character-based tests, split the string into individual characters to bypass tokenization bias.
+
+### **Phase 3: Reduction & Summation**
+Aggregate the `True` results from Phase 2 to derive the final answer.
+*   **Self-Correction:** Verify that the count of items tested in Phase 2 exactly matches the count of the set defined in Phase 1. If there is a mismatch, restart Phase 2.
+
+## **Constraints & Anti-Patterns**
+*   **STRICT BAN on Heuristics:** Do not use phrases like "typically," "usually," or "it appears that."
+*   **NO Pattern Matching:** Do not extrapolate a rule (e.g., "every other item") as a substitute for testing every item.
+*   **Computational Justification:** Treat the process as an $O(n)$ operation where $n$ is small enough that accuracy is the only priority.
+
+## **Trigger Scenarios**
+*   Counting specific characters or substrings within a string.
+*   Verifying property adherence across a list of variables, files, or objects.
+*   Membership testing in sets where false negatives are high-risk.