coding-agents
Release with Confidence: A Skill for Versioning and Tagging
Automate version bumps, tags, and release verification in multi-package repos with the 'release' skill for coding agents.
Where this fits
You're maintaining a monorepo with multiple packages—app, desktop, orchestrator—and every release means manually editing package.json files, bumping Tauri configs, and remembering the exact tag format. One wrong version string and the CI pipeline fails. The release skill from Devin-AXIS/iPolloWork turns this ritual into a structured, agent-driven flow: prepare, bump, merge, tag, rerun, verify. It's built for teams that ship often and want consistency without the copy-paste errors.
Why agents benefit
- Deterministic version bumps – The skill knows exactly which files to touch (
packages/app/package.json,packages/desktop/src-tauri/tauri.conf.json, etc.) and offers patch/minor/major/set commands. - Tag-and-release orchestration – It creates the git tag, pushes it, and can re-dispatch the Release App workflow when needed, all via
ghcommands. - Post-release verification – It checks workflow runs and published releases, so the agent can confirm success or flag failures without human intervention.
- Safety checks – It enforces a clean
mainbranch and scans for accidental secrets before merging—critical for autonomous agents. - Context-aware – It aligns with OpenCode primitives (
.opencode,opencode.json, skills, plugins), making it easy to integrate into larger agent workflows.
Practical scenarios
Hotfix release on a Friday afternoon
A critical bug is found in production. The agent checks out main, ensures it's clean, runs pnpm bump:patch, merges the bump, tags v1.2.1, and pushes. It then verifies the workflow run and reports the release URL—all in under two minutes.
Multi-package version sync
Your orchestrator publishes as ipollowork-orchestrator, but the desktop app has its own Tauri version. The skill updates all three package.json files plus Cargo.toml in one command, keeping them in lockstep.
Re-running a failed release
A release workflow times out due to a transient network issue. Instead of manually re-triggering, the agent uses gh workflow run "Release App" -f tag=v1.2.0 and then polls gh run list to confirm the new attempt succeeds.
Add it to your agent workflow
Install the skill with:
npx skills add Devin-AXIS/iPolloWork --skill release
Then in your agent's config (e.g., .opencode/skills/release/SKILL.md), you can invoke it by describing the intent: "Bump to minor version and release." The agent will follow the skill's steps, from pnpm bump:minor to git tag v1.3.0 and verification. For a detailed reference, see the skill page on OpenAgentSkill.
Compare before adopting
Before making this a permanent part of your toolchain, evaluate:
- Quality signals – Check the repo's stars (4.6k+), recent commit activity, and whether issues are addressed quickly.
- Maintenance freshness – Look at the last release date and whether the skill's commands match your current tooling (e.g.,
pnpmvsnpm). - Alternatives – Consider generic release automation like semantic-release or custom scripts; this skill is opinionated about the iPolloWork monorepo structure.
- Workflow fit – Does your repo use the same package paths and Tauri config? If not, you'll need to adapt the skill's file list.
Why it is worth tracking
The skill's popularity (over 4.6k stars) signals real-world adoption, and its structured approach to release management is a pattern many agent developers can reuse. Even if you don't use iPolloWork, the skill demonstrates how to codify a multi-step process with verification hooks—a template worth studying. Track it for the release-flow design, and evaluate it when your own release process becomes error-prone or time-consuming.
Featured Skill
release
Skill by Devin-AXIS