Repository navigation
release: v2.1.6 — deepwork resilience, fail-fast anti-mocking, and project tool execution - #449
Merged
Merged
Conversation
added 7 commits
November 16, 2026 06:21
…workspace, and rank skills
…lization, and shell cwd fallback
…nd shell cwd binding
ils15
force-pushed
the
fix/deepwork-resilience-and-tool-cwd
branch
from
November 16, 2026 09:26
ca36960 to
de352bf
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
This PR addresses the core framework issues identified during end-to-end multi-agent orchestration runs:
Deepwork Plan Schema & Multi-Decision Normalization:
test,command,file,manual) from string inputs and preserved raw command text.@demeter->demeter) and choices without losing blueprint intent.Tool Execution Context & Directory Fallback:
shellToolto dynamiccreateShellToolusingresolveToolCwd, ensuring command executions without explicit directory target run in the active project workspace instead of falling back to$HOME.pantheon_jev_decideto comply strictly with upstream host validators (eliminating top-leveloneOf/anyOfrestrictions).Skills & TUI Polish:
pantheon_skill_searchwith threshold gating to prevent word-internal substring false positives.Documentation:
README.mddetailing deepwork resilient normalization, tool schema validation, and shell working directory resolution.Verification
npm run lint(biome ci .): Clean (496 files checked).npm run typecheck(tsc --noEmit): Clean.npm run test:quick: 73/73 passed.node --test tests/pantheon/deepwork-plan.test.ts tests/pantheon/tools-shell.test.ts tests/pantheon/jev-decisions.test.ts tests/pantheon/skill-search-tool.test.ts: 30/30 passed.~/pantheon-sandbox/with 303/303 tests passing in 9.8s.