new-target-backend

Add a new target backend skeleton to LLVM 22. Covers CMake registration, TargetInfo, TargetMachine, MCTargetDesc, minimal TableGen register/instruction defs, and wiring into the LLVM build.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./tiles/tessl-llvm/skills/new-target/SKILL.md

Quality

Content

92%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is an excellent, highly actionable skill that provides a complete, copy-paste-ready skeleton for adding an LLVM backend. The workflow is clearly sequenced with verification steps and a useful 'common mistakes' section. The only weakness is that the content is quite long and could benefit from splitting detailed code into referenced sub-files, though the monolithic format is arguably acceptable for a step-by-step tutorial where ordering matters.

Dimension	Reasoning	Score
Conciseness	The skill is lean and efficient throughout. It assumes Claude knows LLVM concepts, C++, CMake, and TableGen. Every section provides only the necessary code and structural information without explaining what LLVM is or how backends work conceptually.	3 / 3
Actionability	Nearly every step contains complete, copy-paste-ready code — full C++ source files, CMakeLists.txt, TableGen definitions, and shell commands. The code is executable and includes proper include guards, namespaces, and registration boilerplate.	3 / 3
Workflow Clarity	The 9-step workflow (Steps 0-8) is clearly sequenced with logical dependencies. Step 8 provides explicit verification commands (cmake build, llc --version \| grep, test codegen), and the 'common mistakes' section serves as a validation checklist for error-prone operations.	3 / 3
Progressive Disclosure	The content is well-structured with clear headers and a logical progression, but it's a long monolithic file (~300+ lines of code) with no references to external files. The directory structure, TableGen defs, MCTargetDesc, and TargetMachine code could be split into referenced sub-documents for better navigation.	2 / 3
	Total	11 / 12 Passed

Description

67%Scale 1-3

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 specificity and distinctiveness, clearly enumerating the concrete components involved in adding an LLVM backend target. Its main weakness is the lack of an explicit 'Use when...' clause, which caps completeness at 2. The trigger terms are domain-appropriate but could benefit from more natural user-facing language variations.

Suggestions

Add an explicit 'Use when...' clause, e.g., 'Use when the user wants to add a new target architecture to LLVM, create a backend skeleton, or port LLVM to a new processor.'

Include natural language trigger variations such as 'new architecture', 'add a target', 'LLVM port', or 'custom backend' to improve discoverability from varied user phrasings.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions: CMake registration, TargetInfo, TargetMachine, MCTargetDesc, TableGen register/instruction defs, and wiring into the LLVM build. These are all concrete, identifiable components.	3 / 3
Completeness	Clearly answers 'what does this do' with specific components, but lacks an explicit 'Use when...' clause or equivalent trigger guidance. The 'when' is only implied by the nature of the task description.	2 / 3
Trigger Term Quality	Includes relevant technical terms like 'LLVM', 'backend', 'TableGen', 'TargetMachine', 'CMake', but these are fairly specialized jargon. Missing more natural user phrases like 'new architecture', 'add a target', 'port LLVM', or 'backend skeleton'. Users might say 'add a new LLVM target' which is partially covered.	2 / 3
Distinctiveness Conflict Risk	Highly distinctive — adding a new LLVM backend target is a very specific niche task. The combination of LLVM 22, backend skeleton, and the specific components (TargetInfo, MCTargetDesc, TableGen) makes it extremely unlikely to conflict with other skills.	3 / 3
	Total	10 / 12 Passed

Validation

100%

Warnings & errors only

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: hbarve1/tessl-llvm
Path: tiles/tessl-llvm/skills/new-target/SKILL.md
Commit: a251bde

Reviewed: 6 days ago

Table of Contents

Discovery Implementation Validation

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.