Evaluate + add caveman-code as a runner agent - #479
Merged
Conversation
Evaluated github.com/JuliusBrussee/caveman-code (npm @juliusbrussee/caveman-code, MIT — license gate passes) live against 0.65.2: binary caveman-code (alias of primary caveman), -p print mode with prompt on stdin, BYOK API-key auth, 20+ providers, install via npm. Implements CodeyBox.Agents.CavemanCode on CliAgentRunnerBase (stdin prompt, --model, allowlisted --thinking), host + in-VM smoke probes, model-list probe over --list-models, quota-failure detector over verified auth/429/overloaded_error shapes, JSON-envelope-only cost extractor, claim-nothing stream parser. Registers AgentKind.CavemanCode with runner/probe/extractor/parser/detector wiring, 8 credential mappings, AgentDefaults + frontier-coding member (openai/gpt-5.5, PayPerApi, score 95, no sensitive tag — frozen upstream), hot-reloadable caveman pricing bucket, install-table row, and agent-quirks section. No IAgentQuotaProbe: BYOK keys expose no quota meter (opencode parity). CodeyBox-Prompt-Revision: 1 Co-Authored-By: CodeyBox <noreply@codeybox.invalid>
EnvironmentCredentialProvider rejects two mappings for the same agent targeting the same sandbox variable, so the eight CavemanCode rows in one provider threw at startup (every host-booting test 500'd). Keep the namespaced CODEYBOX_CAVEMAN_* rows where they were and move the verbatim provider-key fallbacks into the second provider instance, mirroring the existing Codex namespaced/verbatim split. WorkItemDiffEndpoints now passes --git-dir explicitly so bare-repo git diff works on hardened hosts with safe.bareRepository=explicit. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> CodeyBox-Prompt-Revision: 1 Co-Authored-By: CodeyBox <noreply@codeybox.invalid>
DefaultCavemanCodeCliRunner maps IProcessRunner StartFailed to the exit-1/empty result the model-list probe reports as a graceful failure. That runner-to-probe wiring had no test: cover the argv shape plus the mapping, and the end-to-end probe failure through the real runner with only the process boundary stubbed, so a missing caveman-code binary can never surface as a throw or an empty success. CodeyBox-Prompt-Revision: 1 Co-Authored-By: CodeyBox <noreply@codeybox.invalid>
CodeyBox-WorkItem: e6c68eefcef741ceaa4e9edef6e0105d CodeyBox-Agent: copilot/muse-spark-1.3-contributor Co-Authored-By: CodeyBox <noreply@codeybox.invalid>
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.
Automated via CodeyBox — work item e6c68eefcef741ceaa4e9edef6e0105d
Co-Authored-By: CodeyBox noreply@codeybox.invalid
🤖 Generated with CodeyBox