fix: compat issues #57/#58/#59 + Liquid AI tool-call parser (v1.3.1) - #60
Merged
Conversation
#58 zsh shell: always launch bash on POSIX (was honouring $SHELL=zsh while passing bash-only flags --norc, so every bash tool call died with "shell exited"). Also scope shell event handlers to their specific child process to fix a latent race where a superseded shell's late exit/data events stomped a freshly-spawned one. #58 stdin EPIPE: attach a stdin 'error' handler so a dead shell fails the single tool call instead of crashing the whole process. #58 web tools: advertise web_search/web_fetch in the system prompt when SMALLCODE_WEB_BROWSE=true (both prompt builders) so small models stop refusing research tasks. #59 remote Ollama: probe SMALLCODE_BASE_URL (not just OLLAMA_HOST/localhost) for the health check, carry auth, convert URL-embedded basic-auth (https://user:pass@host) into an Authorization: Basic header, and surface the real connection error instead of a generic "not running" message. #57 prebuild-install: documented the harmless upstream deprecation warning in the README (transitive optional dep; no newer version to bump to). Also lands Liquid AI lfm2.x tool-call support: a Python-kwarg parser for the <|tool_call_start|>[func(kw='val')]<|tool_call_end|> format LM Studio passes through verbatim, plus a reasoning_content fallback for empty content. Packaging: bump to 1.3.1; .npmignore now excludes .qartez/.github/test and redundant src/compiled/*.ts sources (173->134 files, 1.4MB->1.1MB unpacked). .gitignore adds .kiro/, .kiro-tmp/, .qartez/, jobs/, bench/harbor/. Tests: +18 cases. Full suite 148 passing.
Doorman11991
force-pushed
the
fix/compat-issues-57-58-59
branch
from
May 29, 2026 08:34
1461a4a to
ae21383
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
#58 zsh shell: always launch bash on POSIX (was honouring $SHELL=zsh while passing bash-only flags --norc, so every bash tool call died with "shell exited"). Also scope shell event handlers to their specific child process to fix a latent race where a superseded shell's late exit/data events stomped a freshly-spawned one.
#58 stdin EPIPE: attach a stdin 'error' handler so a dead shell fails the single tool call instead of crashing the whole process.
#58 web tools: advertise web_search/web_fetch in the system prompt when SMALLCODE_WEB_BROWSE=true (both prompt builders) so small models stop refusing research tasks.
#59 remote Ollama: probe SMALLCODE_BASE_URL (not just OLLAMA_HOST/localhost) for the health check, carry auth, convert URL-embedded basic-auth (https://user:pass@host) into an Authorization: Basic header, and surface the real connection error instead of a generic "not running" message.
#57 prebuild-install: documented the harmless upstream deprecation warning in the README (transitive optional dep; no newer version to bump to).
Also lands Liquid AI lfm2.x tool-call support: a Python-kwarg parser for the <|tool_call_start|>[func(kw='val')]<|tool_call_end|> format LM Studio passes through verbatim, plus a reasoning_content fallback for empty content.
Packaging: bump to 1.3.1; .npmignore now excludes .qartez/.github/test and redundant src/compiled/*.ts sources (173->134 files, 1.4MB->1.1MB unpacked). .gitignore adds .kiro/, .kiro-tmp/, .qartez/, jobs/, bench/harbor/.
Tests: +18 cases. Full suite 148 passing.