modernize
Skill4 starsChanged todayMITView on GitHub
[oh-my-agy] Phased codebase, dependency, framework, and runtime modernization workflow using behavior-lock baselines, license/coverage pre-flight checks, codemod waves, Expand-Migrate-Contract seams, and rollback checkpoints. Use for $modernize or when upgrading packages, frameworks, language versions, SDKs, or legacy modules end-to-end. Flags: --plan, --dry-run, --wave.
- Reads credentials
14 other projects use a version of this.
Adopted by
Most recently taken up by
- cogine-ai/cogine-dev-skillsettheir own version, first seen Nov 16, 2026
- adelrodriguez/skillstheir own version, first seen Nov 15, 2026
- lbussell/agent-skillstheir own version, first seen Nov 8, 2026
- scottwater/skillstheir own version, first seen Nov 8, 2026
- Belweave/hstacktheir own version, first seen Nov 8, 2026
- shubrt/skillstheir own version, first seen Nov 1, 2026
Used with
Projects with it also use takeover, unslop and 1 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
assisjpSkilltakeover
Take over an existing branch, pull request, worktree, or agent handoff while preserving the original goal, contributor credit, and concurrent work. Use when the user asks to take over, finish, rescue, or continue someone else's implementation. Do not use to replace a clear request to start new work.
Also in 4 projects that use this
photon-groveSkillunslop
Detects and rewrites generic, overly polished, or AI-sounding prose while preserving meaning. Use when the user asks to "unslop", humanize, de-AI, tighten, or review whether writing sounds generated.
Also in 3 projects that use this
BlockSkillfrontend-design
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
Also in 3 projects that use this
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- garrytan/gstack★ 142ksame name
- t3dotgg/tstack★ 3.8ksame name
- cogine-ai/cogine-dev-skillset★ 3same name
- mneves75/skills★ 3same name
- connorch/skills★ 2same name
- damacus/agent-skills★ 2same name
- lrstanley/skills★ 2same name
- scottwater/skills★ 2same name
- adelrodriguez/skills★ 1same name
- Belweave/hstack★ 1same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository23 other files it gives its agents.
- rules/AGENTS.mdAGENTS.md
- ai-slop-cleanerSkill
- analyzeSkill
- askSkill
- autopilotSkill
- cancelSkill
- code-reviewSkill
- deep-interviewSkill
- deep-researchSkill
- doctorSkill
- hudSkill
- loftSkill
- performance-goalSkill
- prSkill
- ralplanSkill
- releaseSkill
- skillSkill
- teamSkill
- threat-modelSkill
- ultragoalSkill
- ultraqaSkill
- visual-ralphSkill
- wikiSkill
The file
We split this skill into the rules it gives an agent
11 rules. 0 are common, 3 are distinctive, and 8 are very specific to this project.
0
Common
Most of the 14 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 21
Do NOT invent an upgrade plan blindly: run pre-flight audits (dep-license-diff, coverage-map) and check upstream migration guides …
- Distinctive: line 23
Respect host parity: avoid Antigravity-only scripts or host-locked assumptions when modernizing shared tooling.
- Very specific: line 27
Run read-only diagnostics from the plugin root (or -C <target-repo>) before installing or editing dependencies:
- Very specific: line 29
Supply-chain & license check: run sh <plugin>/hooks/run.sh dep-license-diff <repo> --json (or invoke supply-chain-auditor) to surface …
- Very specific: line 30
Test coverage & risk ranking: run sh <plugin>/hooks/run.sh coverage-map <repo> --json to locate high-LOC/high-export modules lacking test …
- Very specific: line 34
If --plan is set, stop after emitting the wave plan in .omagy/plans/modernize-{slug}.md and confirm scope with the user.
- Very specific: line 40
Wave 0 — Behavior Lock Baseline: add or harden characterization tests via test-engineer covering critical inputs, error paths, …
- Very specific: line 44
Migrate: run targeted codemods or incremental module ports (grep every caller first so shared abstractions update coherently), verifying …
- Distinctive: line 45
Contract: remove temporary adapters and deprecated shims ONLY once 100% of callers pass cleanly under the target version.
- Very specific: line 46
Wave 3 — Docs, Environment & Runtime Parity: run sh <plugin>/hooks/run.sh docs-drift <repo> --json and update README.md, setup guides, …
- Distinctive: line 50
Every wave requires fresh evidence before moving to the next wave:
“Common” is measured: kept by at least 6 in 10 of the 14 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Modernize
Execute controlled, reversible upgrades across dependencies, runtimes, frameworks, and legacy modules without breaking production behavior. Unlike $ralplan --migrate (which stops at planning a migration PRD), $modernize runs pre-flight dependency/coverage audits, locks existing behavior with regression tests, applies incremental codemod or adapter waves, and verifies every checkpoint before advancing.
Usage
$modernize "upgrade Node runtime, dependencies, or framework across target modules"
$modernize --plan "upgrade major ORM and router without writing code yet"
$modernize --wave 1 "bump core packages and lock characterization tests"Boundaries
- Load references/modernization-safety-ladder.md before designing compatibility shims, dual-write seams, or multi-wave rollbacks.
- Do NOT invent an upgrade plan blindly: run pre-flight audits (
dep-license-diff,coverage-map) and check upstream migration guides ($deep-research --quickviaresearcher) first. - Keep waves atomic and reversible: complete characterization tests before editing syntax, API signatures, or lockfiles. Never squash an entire multi-major upgrade into one unverified rewrite.
- Respect host parity: avoid Antigravity-only scripts or host-locked assumptions when modernizing shared tooling.
Pre-Flight Audit & Dependency Triage
Run read-only diagnostics from the plugin root (or -C <target-repo>) before installing or editing dependencies:
- Supply-chain & license check: run
sh <plugin>/hooks/run.sh dep-license-diff <repo> --json(or invokesupply-chain-auditor) to surface unpinned (*,latest), missing lockfiles, or strong copyleft (AGPL, standalone/ANDGPL/SSPL) transitive additions. - Test coverage & risk ranking: run
sh <plugin>/hooks/run.sh coverage-map <repo> --jsonto locate high-LOC/high-export modules lacking test coverage and flag zero-assertion test files before touching them. - Deprecation & breaking-change research: invoke
researcher(orexplorefor internal callers viagrep) to fetch official changelogs, minimum runtime versions, and removed APIs for each target package or SDK. - Inspect
package.json,pyproject.toml,Cargo.toml,go.mod, lockfiles, and.github/workflowsto align CI runtime matrices with the target version.
If --plan is set, stop after emitting the wave plan in .omagy/plans/modernize-{slug}.md and confirm scope with the user.
Step-by-Step Wave Execution
Execute upgrades in strict sequence via executor (or directly in small bounded repos), verifying after every wave:
- Wave 0 — Behavior Lock Baseline: add or harden characterization tests via
test-engineercovering critical inputs, error paths, serialization formats, and public contracts of every un-tested target module reported bycoverage-map. Establish a green baseline before changing dependency versions or code. - Wave 1 — Safe Minor/Patch & Tooling Alignment: apply non-breaking dependency bumps, type definitions, linter/compiler configs, and CI version-matrix updates first. Verify lockfile determinism (
dep-license-diff) and baseline suite pass. - Wave 2 — Expand-Migrate-Contract Seam (Breaking Major / API Upgrades):
- Expand: introduce a compatibility adapter, dual-read/dual-write wrapper, or side-by-side schema column without removing legacy consumers.
- Migrate: run targeted codemods or incremental module ports (
grepevery caller first so shared abstractions update coherently), verifying unit and integration checks per batch. - Contract: remove temporary adapters and deprecated shims ONLY once 100% of callers pass cleanly under the target version.
- Wave 3 — Docs, Environment & Runtime Parity: run
sh <plugin>/hooks/run.sh docs-drift <repo> --jsonand updateREADME.md, setup guides,.env.examplekeys (never inspect.envsecrets), and CI workflows (devops-engineer/doc-steward).
Verification & Rollback Gates
Every wave requires fresh evidence before moving to the next wave:
- Targeted unit/integration suite, typecheck, lint, and build pass with exit code
0. sh <plugin>/hooks/run.sh dep-license-diff <repo> --jsonconfirms zero new copyleft/unpinned regressions.sh <plugin>/hooks/run.sh floor-guard <repo> --jsonreports0suppressed checkers, lost assertions, placeholder code, or hardcoded secrets.- Record each wave's before/after baseline, changed manifests, rollback commit/patch strategy, and residual warnings (e.g. deferred major bumps) before claiming completion. For high-risk runtime shifts, invoke
runtime-qaorverifierfor independent verification.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

