Skip to Content
🎉 Nabla 1.2.0 is out! 🥳 Read the docs →
DocsGlossary

Glossary

This glossary clarifies the terms you’ll see in the CLI, outputs, and policy assessments.

Artifact

An input to analysis. Can be a single binary, an archive, or a directory containing binaries and related files.

Assessment

High‑level evaluation of mapped controls against observed evidence. Produced by nabla assess as JSON (see Outputs & Reporting → Assessment report).

BinaryAnalysis

The structured record Nabla builds for each analyzed artifact (format, arch, symbols, strings, libraries, hashes, sections, metadata). Rules and assessments consume this to produce findings and control statuses.

Confidence (score)

A numeric indicator (0–1) expressing confidence in an assessment outcome or finding, based on signal quality and corroboration. Not a direct measure of business risk.

Config (configuration)

YAML settings that control scanning and policy behavior. Default path: ./.nabla/config.yaml. CLI flags override config values.

Control

An actionable requirement that maps one or more rules to a governance objective. Controls roll up to policies and receive an implementation status in assessments.

Control status

Implementation outcome for a control during assessment: implemented, partially-implemented, or not-applicable.

CWE

Common Weakness Enumeration identifiers attached to relevant findings in SARIF.

CSPRNG

Cryptographically secure pseudorandom number generator (e.g., getrandom, arc4random, CTR_DRBG). Prefer for keys/tokens/nonces.

Evidence

Signals observed during analysis (e.g., symbols, strings, libraries, section flags). Summarized and capped in outputs to keep reports readable.

Evidence tier

Qualifies a finding’s basis: observed (direct evidence) or heuristic (inferred patterns). Heuristic items can be excluded with --exclude-heuristics.

Finding

An individual rule result (e.g., “weak_crypto_algorithms”) with a level, message, properties, and optional locations in SARIF.

Framework

A compliance or best‑practice standard that policies tailor messaging to (e.g., GSMA IoT, ETSI EN 303 645, FIPS 140‑3, NIST SP 800‑53/800‑193, FDA Premarket).

Heuristic

Inference‑based signal that suggests a condition (e.g., likely flash encryption). Marked as heuristic in outputs; can be filtered out.

Help URI

Link from a rule to its documentation (e.g., /docs/rules#weak_crypto_algorithms). Appears in SARIF rule registry and in findings.

Level (severity)

Finding importance used for filtering and exit behavior. Levels: low, medium, high.

Minimum level (--min-level)

Output filter that hides findings below a given level without changing evaluation (useful for concise reports).

Policy

Collection of controls aligned to a governance objective or framework. Policies determine how individual findings roll up to an outcome.

Redaction

Removes sensitive local paths or details from output. Enable with --redact-paths for shareable CI logs.

Remediation

Actionable guidance attached to a finding describing how to resolve the issue.

Rule

Deterministic check that inspects BinaryAnalysis and emits a finding (with severity, message, and properties). See Rules for the registry and details.

SARIF

Static Analysis Results Interchange Format (JSON) for tool‑agnostic reporting. Nabla emits SARIF via --output sarif.

SBOM

Software Bill of Materials. Nabla can generate CycloneDX SBOMs with nabla sbom --from <path|url>.

Signed update

An update mechanism that verifies signatures before applying code or firmware. Related to the signed_update_support rule and update policies.

Secure boot

Boot process that verifies code before execution using a root of trust. Related to secure_boot_present and policy assessments for authenticity.

Strings limit (--strings-limit)

Caps how many embedded strings are collected per file to control performance and output size.

Summary report

Human‑readable overview (Markdown recommended) showing top risks, counts, and links to rule docs. Enable with --summary md.

Timeout (--timeout-ms)

Per‑file analysis time budget in milliseconds. Helps bound worst‑case processing.

TLS

Transport Layer Security. Prefer HTTPS/TLS over plaintext protocols (see insecure_protocol_http).

W^X

Writable‑xor‑Executable memory policy. Findings like writable_executable_section indicate sections that are both writable and executable.

Work concurrency (--concurrency)

Maximum parallel workers used during analysis.

Last updated on