跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
浏览器与自动化

pixee-scan

List, view, analyze, create, and delete Pixee scans with filters for repository, branch, detector tool, and analysis state.

pixee · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add pixee/pixee-cli --skill pixee-scan

技能说明


name: pixee-scan description: "List, view, analyze, create, and delete Pixee scans with filters for repository, branch, detector tool, and analysis state." license: Apache-2.0 compatibility: Requires the pixee CLI binary on PATH metadata: version: 1.0.0 openclaw: category: "developer-tools" requires: bins: - pixee cliHelp: "pixee scan --help"

pixee scan

PREREQUISITES: Read ../pixee-shared/SKILL.md for global flags, exit codes, and error handling, ../pixee-auth/SKILL.md if authentication needs to be configured, and ../pixee-repo/SKILL.md for the --repo resolution protocol. See ../pixee-analysis/SKILL.md for the analysis UUID pixee scan analyze returns, and ../pixee-integration/SKILL.md for the --integration-id discovery flow used by create.

pixee scan manages scans imported into the Pixee platform: list, view, kick off an analysis, upload a new scan, and delete. A scan is the raw output of one detector tool (sonar, appscan, dependabot, datadog_sast, semgrep, codeql, etc.) imported on a specific branch and commit; analyses, findings, and downstream patches all attach back to a scan. Most production scans land via the platform's CI integrations; pixee scan create is the manual import path for bulk-loading historical scans, smoke-testing detector formats, or one-shot uploads. Scans expire roughly seven days after import, and fetching an expired UUID returns a not-found error.

pixee scan list

pixee scan list [filter flags...]

All flags are optional. With none, every scan visible to the token is returned. Pagination is transparent: the CLI walks every page in one call. There is no --paginate flag here; --paginate only lives on pixee api.

Text output is tab-separated with columns id, detector, branch, sha, imported_at. The branch column is empty for detectors that import scans without a branch context (e.g., appscan); JSON output omits the branch field entirely in that case rather than emitting an empty string.

Filter flags:

  • --repo <name-or-uuid> — repeatable. Restrict to one or more repositories. Names resolve via the protocol documented in pixee-repo. Multiple --repo flags OR together; a scan is returned if it matches any of them.
  • --branch <name> — exact branch name (case-sensitive). Works with or without --repo.
  • --tool <name> — repeatable. Filter by detector. Multiple --tool flags OR together.
  • --analysis-state <state> — one of completed, in-progress, not-analyzed. Filters by the state of the scan's downstream analysis pipeline, not the scan import itself.
  • --has-analysis <true|false> — narrow to scans that have at least one analysis associated (true) or none (false).

--analysis-state not-analyzed and --has-analysis false overlap but are not interchangeable: --has-analysis filters on existence, while --analysis-state filters on pipeline state. Pick whichever matches the question being asked.

pixee scan view

pixee scan view <scan-id>

Fetch a single scan by UUID. <scan-id> is the value shown in the first column of pixee scan list.

Default text mode prints a sectioned Key: value block of the scan's headline fields — Id, Detector, Branch, Sha, Imported at, Expires at — colon-separated, one field per line. The Branch line is omitted for detectors that import scans without a branch context (e.g., appscan) rather than emitting an empty value. Use --output json (or --json) for the full HAL representation: _links to the scan's analyses, findings, repository, and scale, plus the body fields id, detector, sha, imported_at, branch (omitted when blank), and expires_at. A non-existent or expired UUID returns the standard not-found error and exits 3.

pixee scan analyze

pixee scan analyze <scan-id> [--finding <finding-id>...] [--watch] [--interval <seconds>]

Start an analysis on a scan. On success the CLI prints Started analysis <analysis-id> on scan <scan-id> and exits 0. The analysis runs asynchronously on the server — capture the UUID from stdout and hand it to pixee analysis view --watch (see pixee-analysis) to poll until terminal state.

Flags:

  • --finding <finding-id> — repeatable. Scope the analysis to one or more specific findings. When omitted, every finding in the scan is analyzed (the common case after an import).
  • --watch — after starting the analysis, poll until it reaches a terminal state. Equivalent to chaining the start call into pixee analysis view --watch on the returned UUID; reach for it when the agent's next step depends on completion.
  • --interval <seconds> — polling cadence with --watch (default 5). Has no effect without --watch.

A scan that's not analyzable (e.g., already deleted, or a scan kind the server cannot run analyses on) returns a 422-shaped problem document and exits non-zero — pixee-shared documents the rendering.

pixee scan create

pixee scan create <repository> --tool <tool> [--file <path>...] [scan metadata flags...]

Upload a scan to a repository. <repository> is name-or-UUID resolved via the protocol in pixee-repo. On success the CLI prints Created scan <scan-id> and exits 0; the upload is multipart, so each --file is attached to the same POST.

Not idempotent. The platform creates a new scan on every successful POST. If a call fails after the upload may have reached the server (timeout, dropped connection), run pixee scan list --repo <id> to check for a freshly-imported scan before retrying.

Flags:

  • --tool <tool> — required. Scanner tool that produced the scan. One of codeql, sonar, semgrep, polaris_sast, polaris_sca, appscan, defectdojo, datadog_sast, dependabot, contrast, gitlab_sast, gitlab_dependency_scanning, snyk, checkmarx, veracode, fortify, arnica_sast. The CLI validates against the enum
阅读完整来源 ↗