Skip to content

chore(deps): update dependency @fission-ai/openspec to v1.14.3 - #1500

Open
renovate[bot] wants to merge 1 commit into
mainfrom
renovate/fission-ai-openspec-1.x
Open

renovate[bot] wants to merge 1 commit into
mainfrom
renovate/fission-ai-openspec-1.x

Conversation

@renovate

@renovate renovate Bot commented Oct 15, 2026 •

Copy link
Copy Markdown
Contributor

This PR contains the following updates:

Package Change Age Confidence
@fission-ai/openspec 1.14.1 → 1.14.3 age confidence

Release Notes

Fission-AI/OpenSpec (@​fission-ai/openspec)

v1.14.3

Compare Source

Patch Changes
  • #​2101 87a69a2 Thanks @​clay-good! - ### Bug Fixes

    • A capability's spec.md may be a symlink to a file elsewhere in the project — list --specs, show, validate and archive no longer refuse openspec/specs/<capability>/spec.md when it links to the file the project keeps somewhere else, such as beside the product documentation it belongs to. A link that resolves outside the project is still refused.
  • #​2103 e33245d Thanks @​surlymochan! - ### Bug Fixes

    • Installing OpenSpec no longer pulls in the unpatched braces package, so a consumer audit does not report GHSA-vfj7-8cjw-p6xm through this package. A generates brace pattern that nests deeply, or expands to more than 1024 paths, is now rejected with a clear message instead of crashing openspec status with a stack overflow or an out-of-memory error.
  • #​2104 276d015 Thanks @​CRYPTONIKAV! - Keep glob artifact paths inside the change directory in the propose, ff-change and continue-change workflows

  • #​2108 5251f2e Thanks @​clay-good! - Stop openspec validate --all hanging on a zero-width TTY. Some hosts hand the
    process a terminal that reports columns === 0 — a pty inherited from a
    pre-commit runner such as lefthook, or script -qec. ora computes how many
    lines to erase as ceil(width / columns), so zero columns makes that count
    Infinity and the clear loop never ends: the command burned a core and wrote
    hundreds of megabytes of cursor escapes without finishing. The spinner is now
    skipped when nothing can be drawn, and validate falls back to its plain output.
    OPEN_SPEC_INTERACTIVE=0 and CI now disable the spinner too, the same way
    --no-interactive already did.

  • #​2110 c9a8ade Thanks @​clay-good! - Preview the built docs site with a first-party static server instead of serve.
    pnpm start in website/ now runs scripts/preview.mjs, a dependency-free
    server that resolves routes in a Next static export (it does not apply
    out/_redirects). This
    drops the serve dev dependency whose transitive tree kept publishing
    advisories (brace-expansion x3, fast-uri, compression) that had to be pinned by
    hand even though none of that code ever reached the published site.

v1.14.2

Compare Source

Patch Changes
  • #​2052 90c95c2 Thanks @​drakeo338! - ### Bug Fixes

    • Archive accepts openspec/, openspec/changes or openspec/specs as a symlink — openspec archive no longer refuses a project-owned symlink that points to another directory, such as a shared docs repository. An archive/ directory that resolves outside the changes directory is still refused.
  • #​2061 7acf282 Thanks @​clay-good! - ### Bug Fixes

    • Guard against deeply nested brace patterns — An artifact generates or apply.tracks pattern that nests braces or parentheses more than 16 levels deep is now rejected with a clear error before glob matching, so a schema cannot overflow the stack through the unpatched braces advisory GHSA-vfj7-8cjw-p6xm.
  • #​2086 9cb5ba8 Thanks @​clay-good! - ### Bug Fixes

    • Explore no longer offers to write default-schema files under a custom schema — The explore workflow's "where to capture" table and example offers now say they describe the default spec-driven schema. With another schema, the agent captures each insight in an artifact that schema defines, names it in the offer, and never creates a file such as design.md that the schema does not have.
  • #​2064 76d49a1 Thanks @​ryandemelo! - Shell Tab completion is faster. openspec __complete, which the completion scripts run on every Tab press, now loads 48 modules instead of 344 when completing changes or specs, and no longer records usage, so a press never waits on the telemetry request and can't use up the first-run telemetry notice where nobody sees it. Completion output is unchanged.

  • #​2083 3fea2a9 Thanks @​clay-good! - The update workflow now states the glob new-file exception at both places that state the scope rule (step 2 and the guardrails), not only where the rule is restated. It said to edit only the concrete files in existingOutputPaths, while the file step 4 creates to fill a gap under a partly written glob artifact is by definition not in that list, so an agent applying the blanket rule refused the operation the flow exists to perform.

  • #​2083 3fea2a9 Thanks @​clay-good! - The verify workflow now reads a REMOVED requirement's baseline from the main spec before searching the codebase for it. A REMOVED delta block carries only the requirement name plus its Reason and Migration notes, so the search started from the heading alone - and because finding nothing is the passing result for a removal, an implementation written in different vocabulary read as a clean removal. When the baseline cannot be recovered from the main spec or that file's history - including a baseline that only restates its own title - Spec Coverage is now reported as not verified for that entry and the search is skipped, instead of passing.

  • #​2054 1bce87f Thanks @​runsonmypc! - openspec validate and openspec archive now warn when a delta spec holds text archive does not carry into the main spec: a ## section other than Purpose and the four delta sections, text above the first ## header, or prose before the first requirement in ADDED or MODIFIED. That text was dropped without notice before. Like other validation warnings, these fail openspec validate --strict. What archive carries is unchanged.

  • #​2063 e0a2f00 Thanks @​gkinter! - Keys in a change's .openspec.yaml that start with a lowercase x- (for example x-goal: G-12 or x-jiraKey) are now treated as your own extension metadata: OpenSpec no longer reports them as unknown, so openspec validate --strict passes on a change that carries them. Any other unrecognized key, including X-Tracker or a bare x-, is still reported.


Configuration

📅 Schedule: (UTC)

  • Branch creation
    • At any time (no schedule defined)
  • Automerge
    • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.


  • If you want to rebase/retry this PR, check this box

This PR was generated by Mend Renovate. View the repository job log.

@renovate
renovate Bot force-pushed the renovate/fission-ai-openspec-1.x branch from 0d0c665 to 9d0db16 Compare October 17, 2026 04:50
@renovate renovate Bot changed the title chore(deps): update dependency @fission-ai/openspec to v1.14.2 chore(deps): update dependency @fission-ai/openspec to v1.14.3 Oct 17, 2026

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants