Skip to content

refactor(pipeline): simplify invoke.json to two paths #807

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
christso merged 1 commit into from
Mar 28, 2026
Merged

refactor(pipeline): simplify invoke.json to two paths #807

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 4 additions & 12 deletions apps/cli/src/commands/pipeline/input.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,8 +130,9 @@ export const evalInputCommand = command({
metadata: test.metadata ?? {},
});

// invoke.json — CLI targets get command info; non-CLI targets get agent
// mode unless subagent_mode_allowed: false forces CLI-based evaluation.
// invoke.json — CLI targets get command info; non-CLI targets get agent mode.
// The manifest carries subagent_mode_allowed so consumers (the bench skill)
// can decide whether to dispatch executor subagents or use `agentv eval`.
if (targetInfo) {
await writeJson(join(testDir, 'invoke.json'), {
kind: 'cli',
Expand All @@ -140,20 +141,11 @@ export const evalInputCommand = command({
timeout_ms: targetInfo.timeoutMs,
env: {},
});
} else if (subagentModeAllowed) {
} else {
await writeJson(join(testDir, 'invoke.json'), {
kind: 'agent',
instructions: 'Execute this task in the current workspace. The agent IS the target.',
});
} else {
// Non-CLI provider with subagent_mode_allowed: false — use agentv eval
// CLI runner instead of executor subagents.
await writeJson(join(testDir, 'invoke.json'), {
kind: targetKind,
subagent_mode_allowed: false,
instructions:
'This target has subagent_mode_allowed: false. Use `agentv eval` CLI to run this target.',
});
}

// criteria.md
Expand Down
Loading