with one click
identify-inconsistencies
Identify inconsistencies under the $1 path
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Menu
Identify inconsistencies under the $1 path
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Address review comments on a pull request -- apply CLAUDE:/SCULPTOR:-prefixed instructions, and critically evaluate feedback from automated reviewers (Vet, Copilot, or any bot)
Prune code comments down to what helps future maintainers. Reviews comments on the current branch's diff and removes incidental history, defensive justification, and correctness arguments. Invoke with /crispy-comments.
End-to-end dev workflow for the minds app stack -- first-time bring-up, every-startup vendor/mngr sync, and the iteration loop against a running Docker agent. Use this when starting or restarting the dev Electron app, or after changing any minds component (mngr, the system interface, the default workspace template).
Create a new PRIVATE GitHub repo that is a full-history copy of imbue-ai/default-workspace-template's current main branch, clone it to <parent-dir>/<repo-name> (default $HOME/project), and push. Use when the user asks to "spin up a new default-workspace-template clone", "fork the default-workspace-template as a private repo", "make me a new private copy of default-workspace-template", or similar.
Use the root justfile as the canonical entry point for ANY minds task -- minds app (desktop client), pool hosts, minds environments (activate/deploy/destroy), minds deployments, and minds tests. Before running ad-hoc `uv run minds ...` / `mngr imbue_cloud ...` commands, check the justfile for a named recipe; if none exists for the task, ADD one. Use whenever the request involves the minds app, pool/leased hosts, a minds env/tier (dev/staging/production), or a minds deploy.
Cut a new production release of the minds app (version bump, FALLBACK_BRANCH, vendor/mngr sync, and minds-v<version> tags on both mngr and default-workspace-template, proven green on CI). The full procedure lives in apps/minds/docs/release.md in the mngr checkout; this skill defers to it. Use when the user asks to "release a new version of minds", "cut a minds release", "bump the minds version", "update the vendored mngr in default-workspace-template", or anything of that shape.
| name | identify-inconsistencies |
| argument-hint | ["target_path"] |
| description | Identify inconsistencies under the $1 path |
The argument $1 is the path to scan. It may be an entire library (e.g. libs/mngr, or just the bare name mngr) or any subdirectory within one (e.g. libs/mngr/imbue/mngr/cli), so you can scope this skill narrowly to part of a library when that is all you care about.
Before doing anything else, resolve these two things from $1 and state them explicitly:
$1 itself, resolved to a real path. (If $1 is a bare library name like mngr, resolve it to libs/mngr or apps/mngr -- whichever exists.) You must only report findings for code under this path.libs/<name> or apps/<name>, so the containing library is exactly that two-component prefix of the scan-scope path (e.g. for a scan scope of libs/mngr/imbue/mngr/cli, the containing library is libs/mngr; for a scan scope of libs/mngr, it is libs/mngr itself). You need this both to gather context and to know where to write the output file -- so make sure you have identified it unambiguously before continuing.Go gather all the context for the containing library (per instructions in CLAUDE.md). Even when the scan scope is a small subdirectory, you still need the whole containing library's context (style guide, primitives, data_types, interfaces, utils) to evaluate the code in context. Be sure to read the containing library's non_issues.md as well.
Once you've gathered that context, please do the below.
Your task is to identify inconsistencies within the scan scope.
In particular, focus on the code, and look for things that are done in different ways in different places, inconsistent variable/function/class naming, and any other code-level inconsistencies.
Do NOT worry about docstrings, comments, or documentation--focus only on the code itself (those will be covered by another task).
Do NOT worry about inconsistencies between the docs/specs and the code either (those will also be covered by another task).
Do NOT report issues that are already covered by an existing FIXME
Do NOT report issues that are highlighted as non-issues in non_issues.md
After reviewing all the code in the scan scope, think carefully about the most important inconsistencies.
Then put them, in order from most important to least important, into a markdown file in the containing library's "_tasks/inconsistencies/" folder (make one if you have to, and always use the containing library's folder even when the scan scope was a subdirectory, so the findings live where create-fixmes and the other identify-* outputs expect them). Name the file ".md` (where you should get "date" by calling this precise command: "date +%Y-%m-%d-%T | tr : -")
For the format of the file, use the following:
# Inconsistencies under <scan scope> (identified on <date>)
## 1. <Short description of inconsistency>
Description: <detailed description of the inconsistency, including file names and line numbers where applicable>
Recommendation: <your recommendation for how to fix the inconsistency>
Decision: Accept
## 2. <Short description of inconsistency>
Description: <detailed description of the inconsistency, including file names and line numbers where applicable>
Recommendation: <your recommendation for how to fix the inconsistency>
Decision: Accept
...
There's no need to commit when you're done (these files are gitignored). Just be sure to create the file in the right location with the right content.