How do I install the Preflight Validator skill?

Run npx skillmds@latest add nguyenthienthanh/preflight-validator in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Preflight Validator safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Preflight Validator?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Preflight Validator free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Preflight Validator?

nguyenthienthanh (@nguyenthienthanh) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Preflight Validator by nguyenthienthanh

Name: Preflight Validator
Author: nguyenthienthanh

Preflight Validator

On-demand wrapper over scripts/preflight/run-all.sh. Runs Tier 1 bash linters against a file, command, or arbitrary content. Returns 0 pass / 1 warn / 2 fail. Used by /aura-frog:preflight check and contributors authoring custom hooks.

File contents

AI-consumed reference. Optimized for Claude to read during execution. Human-readable explanation: see docs/architecture/HIERARCHICAL_PLANNING.md or docs/getting-started/ depending on topic.

Pre-flight Validator

STATUS — v3.7.0-beta.1. Thin wrapper over aura-frog/scripts/preflight/run-all.sh.

When this skill runs vs. when the hook runs

hooks/pre-flight-validate.cjs — auto-fires on every PreToolUse for Bash/Edit/Write/Read. Blocks tool call on fail. You don't have to do anything for this — it's automatic.
This skill — explicit invocation when you want to validate something the hook doesn't see. Examples: validating a markdown file before committing, dry-running a Bash command's safety, batching a directory of plan files through frontmatter validation.

Behavior

Resolve target: file path / Bash command string / stdin content
Run bash aura-frog/scripts/preflight/run-all.sh with the appropriate flags
Surface exit code + stderr to caller

Tier 1 linters (per spec §20.2)

Script	Checks
`validate-frontmatter.sh`	YAML frontmatter on plan/skill/agent/rule/command markdown
`validate-tool-input.sh`	Tool input shape — required fields, absolute paths, no-op edits
`validate-tool-output.sh`	ANSI-escape volume, prompt-injection phrases, JSON-claim sanity
`check-path-safety.sh`	Reject path traversals + system files (`/etc/passwd` etc.) + credential dirs
`check-command-allowlist.sh`	Hard-block destructive (`rm -rf /`, `mkfs`, fork bombs); warn on risky (`git push --force`, `DROP TABLE`)
`check-secret-patterns.sh`	High-confidence credential patterns (AWS, GitHub, OpenAI, JWT, private keys)

run-all.sh orchestrates and returns the highest exit code from any linter.

Exit codes (per spec §20.2)

exit_codes[3]{code,meaning,downstream_action}:
  0,pass,"continue silently"
  1,warn,"emit warning, allow"
  2,fail,"BLOCK tool call (PreToolUse hook returns 2 → claude-code rejects the tool)"

Invocation patterns

Auto — hook on PreToolUse fires this for every tool call
Manual — /aura-frog:preflight check runs run-all.sh against current tool context
CI — bash aura-frog/scripts/preflight/run-all.sh --files <files> for batch validation
Contributor — invoke individual scripts directly when authoring a new hook

What this skill does NOT do

Does NOT run OPA Tier 2 (deferred to v3.7.0-rc.1+; OPA is optional per spec §20.4)
Does NOT mutate tool inputs/outputs (read-only)
Does NOT cache results (each call is fresh — fast enough at ~10ms)
Does NOT fall back when bash is unavailable (Linux/macOS required)

Tie-Ins

Spec: §9.7, §20 (pre-flight)
Script: aura-frog/scripts/preflight/run-all.sh — implementation
Hook: hooks/pre-flight-validate.cjs — primary auto-trigger
Command: /aura-frog:preflight — user-facing entry
Rule: rules/workflow/preflight-policies.md — when, what, bypass policy
Future (rc.1): aura-frog/scripts/preflight/install-opa.sh for Tier 2 OPA policies (optional)

---name: preflight-validatordescription: On-demand wrapper over scripts/preflight/run-all.sh. Runs Tier 1 bash linters against a file, command, or arbitrary content. Returns 0 pass / 1 warn / 2 fail. Used by /aura-frog:preflight check and contributors authoring custom hooks.---> **AI-consumed reference.** Optimized for Claude to read during execution.> Human-readable explanation: see [docs/architecture/HIERARCHICAL_PLANNING.md](../../../docs/architecture/HIERARCHICAL_PLANNING.md)> or [docs/getting-started/](../../../docs/getting-started/) depending on topic.# Pre-flight Validator**STATUS — v3.7.0-beta.1.** Thin wrapper over `aura-frog/scripts/preflight/run-all.sh`.## When this skill runs vs. when the hook runs- **`hooks/pre-flight-validate.cjs`** — auto-fires on every PreToolUse for Bash/Edit/Write/Read. Blocks tool call on fail. **You don't have to do anything for this — it's automatic.**- **This skill** — explicit invocation when you want to validate something the hook doesn't see. Examples: validating a markdown file before committing, dry-running a Bash command's safety, batching a directory of plan files through frontmatter validation.## Behavior1. Resolve target: file path / Bash command string / stdin content2. Run `bash aura-frog/scripts/preflight/run-all.sh` with the appropriate flags3. Surface exit code + stderr to caller## Tier 1 linters (per spec §20.2)| Script | Checks ||---|---|| `validate-frontmatter.sh` | YAML frontmatter on plan/skill/agent/rule/command markdown || `validate-tool-input.sh` | Tool input shape — required fields, absolute paths, no-op edits || `validate-tool-output.sh` | ANSI-escape volume, prompt-injection phrases, JSON-claim sanity || `check-path-safety.sh` | Reject path traversals + system files (`/etc/passwd` etc.) + credential dirs || `check-command-allowlist.sh` | Hard-block destructive (`rm -rf /`, `mkfs`, fork bombs); warn on risky (`git push --force`, `DROP TABLE`) || `check-secret-patterns.sh` | High-confidence credential patterns (AWS, GitHub, OpenAI, JWT, private keys) |`run-all.sh` orchestrates and returns the **highest** exit code from any linter.## Exit codes (per spec §20.2)```toonexit_codes[3]{code,meaning,downstream_action}: 0,pass,"continue silently" 1,warn,"emit warning, allow" 2,fail,"BLOCK tool call (PreToolUse hook returns 2 → claude-code rejects the tool)"```## Invocation patterns- **Auto** — hook on PreToolUse fires this for every tool call- **Manual** — `/aura-frog:preflight check` runs run-all.sh against current tool context- **CI** — `bash aura-frog/scripts/preflight/run-all.sh --files <files>` for batch validation- **Contributor** — invoke individual scripts directly when authoring a new hook## What this skill does NOT do- Does NOT run OPA Tier 2 (deferred to v3.7.0-rc.1+; OPA is optional per spec §20.4)- Does NOT mutate tool inputs/outputs (read-only)- Does NOT cache results (each call is fresh — fast enough at ~10ms)- Does NOT fall back when `bash` is unavailable (Linux/macOS required)## Tie-Ins- **Spec:** §9.7, §20 (pre-flight)- **Script:** `aura-frog/scripts/preflight/run-all.sh` — implementation- **Hook:** `hooks/pre-flight-validate.cjs` — primary auto-trigger- **Command:** `/aura-frog:preflight` — user-facing entry- **Rule:** `rules/workflow/preflight-policies.md` — when, what, bypass policy- **Future (rc.1):** `aura-frog/scripts/preflight/install-opa.sh` for Tier 2 OPA policies (optional)

Frequently asked questions