
Run Scan
- 46 installs
- 126 repo stars
- Updated August 4, 2026
- seqra/opentaint
Helps with ai & agent building tasks.
About
run-scan is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.
- run-scan
- AI & Agent Building
- AI-coding skill
Run Scan by the numbers
- 46 all-time installs (skills.sh)
- +1 installs in the week ending Aug 5, 2026 (Skillselion tracking)
- Ranked #7,629 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
- Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/seqra/opentaint --skill run-scanAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 46 |
|---|---|
| repo stars | ★ 126 |
| Last updated | August 4, 2026 |
| Repository | seqra/opentaint ↗ |
What it does
Helps with ai & agent building tasks.
Files
Skill: Run Scan
Run an OpenTaint scan over a project and collect results
Inputs
From the caller; if omitted, fall back to the default. Ask only when a required input is missing and has no sensible default
- Target
<model-dir>/<project-src>— pre-compiled model or source project directory. Default: model at.opentaint/project - Ruleset
<rules-dir>— Default:builtinplus.opentaint/rulesif present - Rule IDs
<full-id>(optional) — full IDs to restrict the scan to, omit to run all loaded rules - SARIF output
<report.sarif>— Default:.opentaint/results/report.sarif - PassThrough config
<config-dir>(optional) — a passThrough YAML file or a directory of them. Default:.opentaint/pass-through - Dataflow approximations directory
<approx-dir>(optional) — Default:.opentaint/dataflow
Workflow
Point at the code either way: a source project (CLI compiles it) as the positional scan <project-src>, or a pre-built model via --project-model <model-dir>. If project model provided prefer using it instead of source project
opentaint scan --project-model <model-dir> \
-o <report.sarif> \
--ruleset builtin --ruleset <rules-dir> \
--track-external-methodsAppend optional flags as needed:
--rule-id <full-id>— restrict to specific rules (repeatable); omit to run all loaded rules--passthrough-approximations <config-dir>— apply passThrough configs from a YAML file or a directory of them (OVERRIDE: merged with built-ins at the rule level, a provided rule overrides a built-in only when it matches one; repeatable)--dataflow-approximations <approx-dir>— apply code-based approximations (Java sources, auto-compiled; or pre-compiled.classdirs, passed through as-is)
Output
Three files, all next to the SARIF report:
1. <report.sarif> — findings with code-flow traces 2. dropped-external-methods.yaml — methods where dataflow facts were killed (no approximation model) → candidates to approximate; possible source of false negatives 3. approximated-external-methods.yaml — methods already modeled
Key Flags
| Flag | Purpose |
|---|---|
--project-model | Pre-compiled model directory (omit to scan a source project via the positional arg) |
--ruleset | Rule directory (repeatable); builtin for built-ins |
--rule-id | Restrict to specific full rule IDs (repeatable) |
--passthrough-approximations | passThrough configs: a YAML file or directory of them (OVERRIDE, repeatable) |
--dataflow-approximations | Directory of Java sources or compiled classes (repeatable) |
--track-external-methods | Emit dropped-external-methods.yaml + approximated-external-methods.yaml next to the SARIF |
--timeout | Analysis timeout (default 900s) |
Gotchas
- Paths fall back to the
.opentaint/layout when the caller omits them; the caller can override any of them - Duplicate approximation targeting the same class as a built-in errors out