Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/buildmoonshot/skillpacks/refactor-safelynpx skills add buildmoonshot/skillpacks --skill refactor-safelygit clone --depth 1 https://github.com/buildmoonshot/skillpacksWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00066 | $0.00443 |
| Opus 5 | $0.00033 | $0.00221 |
| Sonnet 5 | $0.00013 | $0.00089 |
| Haiku 4.5 | $0.00007 | $0.00044 |
Grade A, and why
refactor-safely scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Refactor Safely
A refactor changes structure, not behavior. The whole risk is changing behavior by accident. Defend against it.
The procedure
-
Pin the behavior first. Before touching anything, make sure the code you're about to move is covered by tests. If it isn't, write characterization tests that capture what it currently does (even quirks) and get them green. You cannot refactor safely what you cannot observe.
-
Refactor in small steps. Make one structural change at a time — extract a function, rename a symbol, move a file. Keep each step mechanical and reversible.
-
Run the tests after every step. Green after each step means the behavior held. If a step goes red, you know exactly which change broke it — revert just that step.
-
Keep refactor commits separate from behavior changes. Never mix "I reorganized this" with "I also fixed a bug / changed logic" in the same diff. A reviewer should be able to trust a refactor diff changes nothing functional. If you spot a real bug mid-refactor, note it and handle it in its own change.
Signs you've left "refactor" and entered "rewrite"
- The tests need to change to keep passing → you're altering behavior, not refactoring. Stop and make that an explicit, separate decision.
- The diff is large and you can't point to the equivalent old code for each new line → break it into smaller steps.
Why this matters
"Just cleaning this up" is one of the most common ways working software breaks, because the change feels safe and goes unreviewed. Characterization tests plus small, individually-verified steps turn a risky sweep into a sequence of provably-behavior-preserving moves.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 28 lines · 66 tokens per session scan A 91c9d8910102
refactor-safely is a skill published in the GitHub repository buildmoonshot/skillpacks (2 stars, last pushed 2mo ago), licensed MIT. It adds 66 tokens to every session and 443 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
bridge
Use when the user wants hyperflow's behavioral rules to apply outside the terminal CLI — in Claude Code Desktop, claude.ai web, or IDE extensions that don't load CLI plugins. Writes a managed doctrine block into the project's CLAUDE.md so autonomy + intent-routing + commit cadence + role separation + file-first rules…
fable-method
A step-by-step problem-solving loop (classify the ask, define done, gather evidence, decide, act surgically, verify by observation, report outcome-first). Use when the user says "/fable-method", "use the fable method", or "approach this like Fable", or proactively when starting any multi-step task that no…
fable-domain
Discuss a domain with the user, research it from real sources, then generate a trusted skill bundle for it - a step-by-step workflow with a flowchart, a domain adapter, a trap fixture, and a smoke eval. Use when the user says "/fable-domain ", "make a skill for ", "add a domain to the fable method", or "give a lesser…
writers-pattern
Add a new platform writer module in src/writers/ that generates and writes agent config files for a supported platform. Each writer exports a function that accepts a config interface, creates directories (rules/, skills/, mcp configs), writes files with proper formatting and frontmatter, and returns string[] of…
caliber-testing
Writes Vitest tests following project patterns: tests/ directories, vi.mock() for module mocking with vi.hoisted() for test-time factories, global LLM mock from src/test/setup.ts, environment variable save/restore in beforeEach/afterEach, vi.clearAllMocks() lifecycle, and test file organization. Use when user says…
llm-provider
Adds a new LLM provider implementing LLMProvider interface with call() and stream() methods. Integrates with provider factory in src/llm/index.ts, config detection in src/llm/config.ts, and error handling via tracking and recovery. Use when adding a new model backend, integrating a third-party LLM API, or extending…