- Bump mthds dependency to >=0.1.1.

[lchoquel]

Note

Medium Risk
Medium risk because several CLI commands now force needs_model_specs=True, which may introduce new network/runtime dependencies and change validation/inputs generation behavior. The pipelex run output printing path is also adjusted to handle missing main_stuff, which could affect CLI output expectations.

Overview
Bumps the package to v0.20.7 and updates the mthds dependency to >=0.1.1 (with matching changelog entry).

Changes multiple non-inference CLI entry points (including agent inputs commands and pipelex build inputs/output, show pipe, and which) to initialize Pipelex with needs_model_specs=True so they load real gateway model specs without enabling inference.

Makes pipelex run only pretty-print the final output when main_stuff is present by adding PipeOutput.optional_main_stuff, and updates VS Code debug launch args to the newer run bundle/run pipe forms.

^{Written by Cursor Bugbot for commit b6f5b7a. This will update automatically on new commits. Configure here.}

[cubic-dev-ai]

No issues found across 3 files

[cubic-dev-ai]

2 issues found across 7 files (changes from recent commits).

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="pipelex/cli/commands/show_cmd.py">

<violation number="1" location="pipelex/cli/commands/show_cmd.py:233">
P1: Don't load real model specs in `show pipe`; this makes a local inspection command fail on gateway/remote-config setup.

(Based on your team's feedback about keeping read-only CLI commands from depending on inference setup.) [FEEDBACK_USED]</violation>
</file>

<file name="pipelex/cli/commands/which_cmd.py">

<violation number="1" location="pipelex/cli/commands/which_cmd.py:85">
P2: `pipelex which` should stay local-only; `needs_model_specs=True` now makes it depend on gateway terms/remote config.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}