v4.27.0

What's Changed

• feat: bundle all skills in CLI + single agentv-dev wrapper + subagent pipeline by @tsoyangbot in #1231

New Contributors

• @tsoyangbot made their first contribution in #1231

Full Changelog: v4.26.1...v4.27.0

v4.27.0-next.1

What's Changed

• feat: bundle all skills in CLI + single agentv-dev wrapper + subagent pipeline by @tsoyangbot in #1231

New Contributors

• @tsoyangbot made their first contribution in #1231

Full Changelog: v4.26.1...v4.27.0-next.1

v4.26.1

What's Changed

• refactor(core): migrate LLM provider layer from Vercel AI SDK to pi-ai by @christso in #1206
• feat: code-grader plain-text fallback + workspace env preflight by @christso in #1209
• fix(core): derive score from assertions when score absent in code-grader by @christso in #1212
• fix(cli): detect global npm self-update installs by @christso in #1213
• fix(core): drop api_format from Azure targets; surface pi-ai errors by @christso in #1214
• fix(cli): use next dist-tag for self-update when on a prerelease version by @christso in #1215
• feat(studio): show DD/MM HH:mm · target · pass% run label with time-ago by @christso in #1218
• feat(cli): polished eval resumability — wizard entry, docs, no-args fallthrough by @christso in #1217
• feat(studio): expose eval resumability — API + Resume action on run detail by @christso in #1220
• fix(cli): resolveCliPath off-by-one in src + dist fallback paths by @christso in #1223
• fix(studio): show target in sidebar label for active/in-progress runs by @christso in #1225
• feat(studio): pass/fail icon in eval detail h1, remove duplicate header by @christso in #1227
• chore(studio): rename Re-run with Filters to Run evals by @christso in #1230
• feat(studio): Stop run button + graceful CLI interrupt by @christso in #1228

Full Changelog: v4.25.1...v4.26.1

v4.26.1-next.1

What's Changed

• fix(studio): show target in sidebar label for active/in-progress runs by @christso in #1225
• feat(studio): pass/fail icon in eval detail h1, remove duplicate header by @christso in #1227
• chore(studio): rename Re-run with Filters to Run evals by @christso in #1230
• feat(studio): Stop run button + graceful CLI interrupt by @christso in #1228

Full Changelog: v4.26.0-next.1...v4.26.1-next.1

v4.26.0-next.1

What's Changed

• feat(studio): expose eval resumability — API + Resume action on run detail by @christso in #1220
• fix(cli): resolveCliPath off-by-one in src + dist fallback paths by @christso in #1223

Full Changelog: v4.25.6-next.1...v4.26.0-next.1

v4.25.6-next.1

What's Changed

• feat(studio): show DD/MM HH:mm · target · pass% run label with time-ago by @christso in #1218
• feat(cli): polished eval resumability — wizard entry, docs, no-args fallthrough by @christso in #1217

Full Changelog: v4.25.5-next.1...v4.25.6-next.1

v4.25.5-next.1

What's Changed

• fix(cli): use next dist-tag for self-update when on a prerelease version by @christso in #1215

Full Changelog: v4.25.4-next.1...v4.25.5-next.1

v4.25.4-next.1

What's Changed

• fix(core): drop api_format from Azure targets; surface pi-ai errors by @christso in #1214

Full Changelog: v4.25.3-next.1...v4.25.4-next.1

v4.25.3-next.1

What's Changed

• fix(core): derive score from assertions when score absent in code-grader by @christso in #1212
• fix(cli): detect global npm self-update installs by @christso in #1213

Full Changelog: v4.25.2-next.1...v4.25.3-next.1

v4.25.2-next.1

What's Changed

• refactor(core): migrate LLM provider layer from Vercel AI SDK to pi-ai by @christso in #1206
• feat: code-grader plain-text fallback + workspace env preflight by @christso in #1209

Full Changelog: v4.25.1...v4.25.2-next.1