@code-partner/codepipe 0.14.1-dev.342.g5ef1cc1d → 0.14.1-dev.346.g99807610

package/dist/skills/container-tooling/SKILL.md

@@ -42,9 +42,15 @@ wastes a whole job on a false result.
 - **No network for the workload by default is not the case** — the container has
   network. Do not rely on isolation as a safety property here.
 - **A service you need running** (a database for a test suite) goes up with
-  `-d` and a published port, and must be stopped when you are done:
-  `podman run -d --rm --name db -p 5432:5432 postgres:16`, then
-  `podman rm -f db`. A leaked container costs the next job on this host.
+  `-d` and a published port, and must be stopped when you are done. Other
+  jobs may run on this host at the same time, so NEVER use a fixed container
+  name or a fixed host port — derive the name from your workspace and let the
+  host port be assigned:
+  `NAME="db-$(basename "$PWD")-$$"; podman run -d --rm --name "$NAME" -p 127.0.0.1::5432 postgres:16`,
+  find the assigned port with `podman port "$NAME" 5432`, and stop it with
+  `podman rm -f "$NAME"` — only ever by that name, never by image or a
+  guessed name (you would kill another job's service). A leaked container
+  costs the next job on this host.
 - **Pull failures are usually rate limits**, not typos. Say so plainly in your
   result rather than retrying in a loop.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@code-partner/codepipe",
-  "version": "0.14.1-dev.342.g5ef1cc1d",
+  "version": "0.14.1-dev.346.g99807610",
   "description": "CodePipe turns tasks into reviewed pull requests: an AI agent analyzes, implements, and opens a PR under your control.",
   "type": "module",
   "main": "./dist/index.js",