v1.5.2

Changelog

• 1fe10b0 Prepare v1.5.2 release
• 8864618 Merge pull request #46 from agent-ecosystem/fix/head-fallback-to-get
• 5d0e862 fix: head 404/405 falls back to GET

v1.5.1

Changelog

• 9eaa319 Prepare v1.5.1 release
• 9c79ec7 Add security reporting info
• 4461508 Limit path traversal to skill dir
• 44b1343 Improve link checker handling of unsafe links
• ce59d79 Add the normal open source contributing artifacts

v1.5.0

Changelog

• 2391cd2 Prepare v1.5.0 release
• eb0e3af Merge pull request #44 from agent-ecosystem/fix/claude-cli-isolation
• 9bfbb79 Remove '--bare' flag because it also strips auth
• e619cb1 fix: add preflight check and --bare for claude CLI client
• 225a296 Merge pull request #43 from marc0olo/feat/claude-cli-provider
• f4fa293 feat: add claude-cli provider for LLM scoring without API keys

v1.4.0

Changelog

• ba40a0b Prepare v1.4.0 release
• 332b52c Merge pull request #42 from agent-ecosystem/refactor-string-flag-handling
• a0d5434 refactor: flag string splitting to use 'StringSliceVar' for consistency and correctness
• 8b3f2ae Merge pull request #41 from agent-ecosystem/feat-allow-dirs
• ee042e7 Add unstaged test fixture to fix failing tests
• ea7923e Add support for explicitly allowing specified directories in a skill bundle
• cc5d9d7 Add a link to the Agent Skill Implementation project

v1.3.1

Changelog

• 6b1d50b Prep v1.3.1 release
• 76b762e Add OWNERS.yaml and OWNERS to known extraneous files for more helpful reporting
• 5ac873c Support cancellation through context, let first call through free
• e865b72 Merge pull request #34 from x3c3/perf/cache-token-encoder
• cee833f Merge pull request #37 from x3c3/feat/rate-limit-llm-calls
• f304631 Merge pull request #35 from x3c3/refactor/deduplicate-regex
• 49f753f feat: add rate limiting for LLM API calls
• 39f986a refactor: deduplicate regex patterns into util/regex.go
• edc157d perf: cache token encoder with sync.Once

v1.3.0

Changelog

• 4eced97 Prepare for v1.3.0 release
• d2b4a84 Merge pull request #28 from agent-ecosystem/develop-1.3.0
• 76f4b62 Rename to '--allow-flat-layouts' for consistency/Go idiomaticity
• f63c195 Add an optional '--accept-flat-layouts' flag
• 843fd9d Add '--allow-extra-frontmatter' flag
• 7a1af3f Update CHANGELOG for v1.2.1

v1.2.1

Changelog

• 25975fc Add sentence detection to keyword stuffing heuristic
• 2bc4d92 Add sponsor button

v1.2.0

Changelog

• e288028 Bump version constant for minor release
• 3edaa8c Add CI and review-skill examples
• 87e9286 Bump default OpenAI model to GPT 5.2

v1.1.0

Changelog

• 5466337 Bump minor version for model name truncation increase
• 78949b8 Increase model name truncation limit in eval compare report

v1.0.0

Changelog

• 0329f8d Use a GitHub runner for CI
• 59d973e Update workflows to run on self-hosted runner
• 2a78caf Move repo from dacharyc to agent-ecosystem
• cb4476b Remove language referring to moved code
• 84f9e5b Add a changelog
• aba6e84 Define stability policy
• bc19ef6 Add more comprehensive lib docs
• 95b773a Add proper godoc comments
• c30aa2e Merge pull request #21 from dacharyc/validation-lib-refactor
• f221067 Final cleanup pass for lib/CLI separation
• 8053c5e Bugfixes, splitting link functionality, and other cleanup
• 70c35fa Rename redundant method names
• 0926dc5 Split evaluate into scoring + presentation
• ac156bd Extract orchestration from 'cmd' to Library
• 03b2eec Judge API surface cleanup
• e3d89ce Factor out shared types and utils
• cc31038 Fix lint errors
• f5e22a0 Break out print functionality to lib
• 9d33f88 Mock judge.LLMClient to improve test coverage
• 696bdb1 Create a new lib package to coordinate evaluating
• 7cab3ff Break out validation, judging, and structural checks into lib