@chankov/agent-skills 0.2.0 → 0.3.0

package/.versions/0.3.0/.pi/extensions/package.json

@@ -0,0 +1,9 @@
+{
+  "name": "agent-skills-pi-extensions-runtime-deps",
+  "private": true,
+  "type": "module",
+  "dependencies": {
+    "@modelcontextprotocol/sdk": "^1.29.0",
+    "typebox": "^1.1.33"
+  }
+}

package/.versions/0.3.0/.pi/harnesses/agent-chain/README.md

@@ -0,0 +1,37 @@
+# agent-chain
+
+Sequential pipeline orchestrator.
+
+> Ported from [`pi-vs-claude-code`](https://github.com/disler/pi-vs-claude-code) by [disler](https://github.com/disler) (MIT). See the [extension catalog](../../../docs/pi-extensions.md).
+
+## What it does
+
+Runs opinionated, repeatable agent workflows. Chains are sequences of agent steps with
+prompt templates: the user's prompt flows into step 1, that output becomes `$INPUT` for
+step 2's template, and so on. `$ORIGINAL` always holds the user's original prompt.
+
+The primary pi agent has **no codebase tools** — it can only kick off a pipeline via the
+`run_chain` tool. On boot you select a chain; the agent decides when to run it. Agents
+keep their session context within a pi session, so re-running a chain resumes each step.
+
+## Commands & tools
+
+- `/chain` — switch the active chain
+- `/chain-list` — list all available chains
+- `run_chain` tool — the only tool the primary agent has
+
+## Requires
+
+- `.pi/agents/agent-chain.yaml` — chain definitions
+- `.pi/agents/*.md` — the agent personas referenced by each chain step
+
+## Usage
+
+```bash
+pi -e .pi/harnesses/agent-chain/index.ts
+```
+
+## Upstream changes
+
+- Theme integration removed — the `themeMap.ts` import and the `applyExtensionDefaults()`
+  call were stripped (this repo does not ship pi themes).