bun-ffi
This skill should be used when the user asks about "bun:ffi", "foreign function interface", "calling C from Bun", "native libraries", "dlopen", "shared libraries", "calling native code", or integrating C/C++ libraries with Bun.
56
63%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./plugins/bun/skills/bun-ffi/SKILL.mdQuality
Discovery
62%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description excels at trigger term coverage and distinctiveness, providing a comprehensive list of natural keywords users would use. However, it completely lacks any description of what the skill actually does — it only says when to use it, not what capabilities it provides. Adding concrete actions (e.g., 'Generates FFI bindings, loads shared libraries via dlopen, maps C types to JavaScript') would significantly improve it.
Suggestions
Add concrete capability descriptions before the trigger clause, e.g., 'Generates FFI bindings for Bun, loads shared libraries via dlopen, maps C data types to JavaScript types, and calls native C/C++ functions from Bun.'
Restructure to lead with 'what it does' followed by 'Use when...' to clearly answer both questions, e.g., 'Provides guidance on using bun:ffi to call native C/C++ functions, load shared libraries, and work with pointers and callbacks. Use when...'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description does not describe any concrete actions or capabilities. It only states when the skill should be used, but never explains what it actually does (e.g., 'generates FFI bindings', 'loads shared libraries', 'calls C functions from Bun'). | 1 / 3 |
Completeness | The 'when' is explicitly and thoroughly answered with multiple trigger scenarios, but the 'what' (what the skill actually does) is entirely missing. The description only specifies triggers without describing capabilities. | 2 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms: 'bun:ffi', 'foreign function interface', 'calling C from Bun', 'native libraries', 'dlopen', 'shared libraries', 'calling native code', and 'C/C++ libraries with Bun' are all terms users would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | The description targets a very specific niche — Bun's FFI for calling native C/C++ code — with distinct trigger terms like 'bun:ffi', 'dlopen', and 'calling C from Bun' that are unlikely to conflict with other skills. | 3 / 3 |
Total | 9 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, example-rich FFI skill with strong actionability — nearly every concept is demonstrated with executable TypeScript code. Its main weaknesses are the lack of error-handling/validation workflows for inherently risky pointer and memory operations, and some verbosity in type listings that Claude already understands. The referenced bundle files don't exist, undermining the progressive disclosure strategy.
Suggestions
Add explicit validation steps after dlopen (check for null/errors), after sqlite3_open (check return code), and after pointer operations to improve workflow safety for these inherently risky FFI operations.
Move the full FFI types table to a `references/type-mappings.md` file (which is already referenced but doesn't exist) and keep only the most common types inline.
Provide the referenced bundle files (`references/type-mappings.md`, `references/performance.md`) or remove the references to avoid broken navigation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The type mapping table is somewhat verbose (Claude already knows C type equivalents), and some sections like Thread Safety and Memory Management are thin enough to be condensed. However, most code examples are lean and purposeful. | 2 / 3 |
Actionability | The skill provides fully executable, copy-paste ready code examples throughout — from basic dlopen usage to structs, callbacks, and a complete SQLite example. Function signatures, type definitions, and calling conventions are all concrete. | 3 / 3 |
Workflow Clarity | The skill covers individual operations well but lacks explicit validation checkpoints. For FFI work involving memory management and pointer operations (destructive/risky), there are no feedback loops for error recovery — e.g., checking return codes from dlopen or sqlite3_open, or validating pointer validity before use. | 2 / 3 |
Progressive Disclosure | The skill references `references/type-mappings.md` and `references/performance.md` at the end with clear loading criteria, which is good. However, no bundle files are provided, so these references are broken. The main file is also quite long (~200+ lines) with sections like the full FFI types table that could be offloaded to a reference file. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- secondsky/claude-skills
- Commit
5e92b71
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.