Coding agents need workflow memory more than more tools
often fail even when they already have the basic tools they need. They can use the shell, git, a browser, and file reading, but they may still choose the wrong process for the job. A release, a hotfix, a , and a are not the same kind of task.
Each one has its own checks, steps that should not be skipped, follow-up updates, and definition of done. Putting all of those rules into the can make the instructions long and messy. A better pattern is to load workflow context only when it is needed.
If the task looks like a release, load the release checklist; if packaging files are being changed, load packaging notes; if it is a , load backup and rules; if it is a hotfix, load and sync rules. When the workflow ends, remove that extra context so the agent is not carrying unnecessary instructions.
Key points
- The issue is often missing workflow memory, not missing tools.
- Releases, hotfixes, s, and s each need different rules.
- A can become too long and confusing if it contains every checklist.
- Loading workflow context only when needed can reduce clutter and token use.
- This can make agents cheaper to run and less likely to skip important process steps.