CompTIA SY0-701: SLSA Supply Chain Levels

A software package can be compiled from reviewed source code and still be compromised if an attacker changes its build environment or published artifact. Signed release files help detect some alterations, but a signature alone cannot explain which source revision and workflow produced the binary. Supply-chain Levels for Software Artifacts (SLSA) describes progressively stronger evidence about software production. Its approved version 1.2 separates Build Track protections from Source Track protections, with different levels and requirements. Understanding those boundaries helps teams design release gates, interpret attestations and avoid misleading claims that one “SLSA badge” makes a product secure.

The supply-chain question is whether the artifact came from the expected process

An attacker might inject a malicious dependency into source control, steal a CI signing token, modify a build script, compromise a hosted runner, or replace a completed binary in a distribution system. These attacks occur at different phases. Source review can catch some malicious changes but not necessarily a compromised build worker. Artifact signing can detect changes after signing but may faithfully authenticate software already compromised before the signature was created.

SLSA provides a framework for expressing increasingly trustworthy information about how software artifacts and source revisions were produced. Build provenance records which builder produced an artifact, which inputs or source revision it used and which process ran. A consumer can compare that record with approved expectations. The goal is not merely to create another JSON file; it is to make an unauthorized build or substituted release harder to pass off as legitimate.

The approved 1.2 specification organizes requirements by tracks. A Build Track level is a claim about the build process and its provenance; a Source Track level is a separate claim about how source revisions are created and protected. A project may have strong hosted build controls but weak review of source changes, or a carefully governed repository with an unreliable artifact publisher. That distinction must survive executive summaries and security questionnaires.

The inventory of software ingredients is useful but solves a different problem. A software bill of materials can enumerate libraries and dependency relationships; SLSA provenance describes where and how a build originated. Combining the two improves release transparency without guaranteeing that a listed dependency is safe or that a signed attestation proves the software’s business logic is correct.

Build substitution example. A developer approves source revision abc123. A compromised publishing job uploads a binary built from a different revision. The package may have a plausible filename and version, but a properly authenticated provenance statement should reveal the unexpected source or builder—and a consumer policy must actually reject that discrepancy.

Build L0 through L3 strengthen different parts of the evidence chain

The approved SLSA 1.2 Build Track uses levels that represent progressively stronger protection of provenance and the build platform. Build L0 states no requirements. Build L1 requires automatically generated provenance showing how an artifact was built; it improves traceability but does not itself provide meaningful tamper resistance. Build L2 requires the artifact to be produced on a hosted build platform that generates and authenticates its provenance, offering stronger protection against alteration after the build.

Build L3 goes further by requiring a hardened build platform designed to resist tampering during the build. This usually involves isolation of build executions, controlled entry points, restricted privileged interfaces and evidence generated by the trusted platform rather than arbitrary build scripts. The exact approved requirements matter; placing a “L3” label on a self-hosted build worker because it runs inside a container is not the same as assessing it against the specification.

Build level Core assurance Important limitation
L0 No formal SLSA build guarantees Do not treat a package as provenance-verified.
L1 Provenance records the build process and inputs Provenance may be easy to forge or alter.
L2 Hosted platform creates authenticated provenance Build-platform compromise is still a significant concern.
L3 Hardened platform resists build-time tampering Does not prove source code is secure or dependencies are benign.

A level describes evidence about a defined threat model, not an overall security grade for an application. A Build L3 project can still ship a vulnerable library or implement broken authorization checks. Treat provenance as a supply-chain control among code review, dependency management, signing, vulnerability testing and runtime monitoring. Supply-chain risk testing should validate the expectations, including whether the artifact actually matches the attested digest.

The Source Track protects how revisions enter a repository

SLSA v1.2 also introduces an approved Source Track with its own requirements. The Source Track is not a footnote added to the Build L3 label: it addresses a different question, whether a source revision was created under trustworthy and enforceable source-control processes. A malicious revision can be built faithfully on a hardened builder, resulting in authentic provenance for an intentionally compromised release. Source protection matters because build assurance cannot reliably infer developer intent.

Source L1 covers version-controlled revisions, enabling precise references to what was changed. L2 adds preserved change history and source provenance. L3 requires enforcement of technical controls defined by the source owner, while L4 requires code review. The exact trust in any revision depends on which controls were in effect when it was created, not simply on whether the repository uses a popular hosting platform. Retrofitting approvals after a compromise does not make an old revision protected retroactively.

For example, a protected branch may require authenticated contributors, review of proposed changes and controls preventing one actor from bypassing organizational policy. A source attestation can help consumers evaluate whether the expected process was followed. But a reviewer can approve a vulnerable change or fail to notice a malicious design; source governance improves traceability and resistance to unauthorized changes without delivering a proof of correct application behavior.

Organizations should declare Build and Source levels separately and state the relevant repositories and artifacts. An SBOM, a repository review record and build provenance are independent forms of evidence. Consumers should test how they connect before relying on one category’s maturity claim to infer completeness in the others. A single chart with a “SLSA 3” label can hide significant differences in coverage and is a poor substitute for a track-specific claim.

Provenance must be authenticated and bound to the artifact being used

Build provenance can describe a trusted platform while still being irrelevant to the binary downloaded by the operator. The consumer needs to bind the attestation to the actual artifact digest, identify the builder, check the cited source revision and evaluate whether the build procedure and inputs conform to policy. This prevents an attacker from supplying an authentic attestation from a different release as an excuse to run an unverified package.

Authenticity matters at two levels. First, the verifier should establish that the statement came from a recognized build service or source-control authority using appropriate signatures and trust configuration. Second, it must assess the claims inside the statement. A correctly signed assertion saying that an unknown runner built the software from a developer’s fork should not pass merely because the signature is valid. The signature protects the statement’s origin and integrity; consumer policy decides whether its content is acceptable.

Artifact identifiers should be stable. A mutable container tag such as “latest” can point to different images on different days. A cryptographic digest identifies the precise artifact bytes and makes comparison with provenance practical. The same discipline supports container image provenance, which combines image identity, attestation verification and the conditions required to deploy an approved artifact. Linking a mutable tag to an old attestation is a common failure mode.

Attestation records need their own security controls. Build services that can issue provenance must protect signing identities and prevent ordinary build jobs from forging declarations about other jobs. If the attacker already controls the trusted platform’s authority, provenance may no longer provide the expected assurance. Threat modeling must include key handling, platform administration, update workflows and a mechanism to remove or distrust a compromised builder.

Release policy turns build evidence into an actual security control

Generating a provenance file without using it in release or deployment decisions is primarily an inventory exercise. A meaningful policy defines which builders, repositories, branches, source review conditions and artifact digests are acceptable for a particular environment. A production admission control can reject a container image without authenticated provenance or with a source revision outside an approved repository. A development sandbox might permit weaker evidence under a clearly limited exception.

Define the behavior when verification is unavailable. A verifier may be unable to reach an attestation store during an outage. Silent allow-all fallback undermines the production guarantee, but a rigid stop with no restoration path can create substantial operational risk. Decide which workloads truly require fail-closed verification, how to restore the service, and who may authorize a time-limited exception. Avoid using one global policy without considering different system availability and safety needs.

Record denial outcomes, not only successful deployments. The policy should distinguish a missing attestation, a malformed document, an untrusted signer, a mismatched digest and an unauthorized source revision. Each condition suggests a different response. For example, a digest mismatch can indicate release substitution or a packaging error, whereas missing metadata may reflect an unsupported legacy pipeline. Observability should make those differences visible without exposing privileged signing secrets.

Verify the rule through negative tests. Attempt to deploy an artifact with correct content but an unrecognized builder, one with an authentic attestation for another digest, and one built from a disallowed branch. Each should produce the expected disposition and evidence. A dashboard showing a high percentage of signed artifacts does not prove that production refuses unauthorized builds.

Consumer expectations must survive vendor handoffs

An enterprise rarely builds every package it deploys. Open-source projects, commercial vendors and internal platform teams may each produce different attestations or use separate build systems. Procurement and security teams should establish what evidence they require, how it will be delivered, which trusted identities can issue it, and when it must be refreshed. An attestation for a generic product family is not necessarily sufficient to identify an installed vendor update.

Ask a supplier whether the specific deliverable can be mapped to a digest and verified provenance, which build service produced it, and whether the supplier has documented accepted source-control processes. The same package can be distributed through different channels; verify that the digest actually received through the selected channel is covered. A well-intentioned distribution mirror can create a gap when files are repackaged without a corresponding trustworthy record.

Where an older component lacks provenance, record the limitation and choose proportionate controls. These might include a controlled software repository, vendor identity verification, independent malware and vulnerability inspection, restricted runtime permissions and an approved migration plan. Declaring every legacy component “SLSA L0, therefore unsafe” does not provide a useful treatment; neither does pretending those components meet higher assurance because a reseller assures customers they follow best practices.

Supplier and dependency assurance still includes security history, patch responsiveness, privileged access, incident notification and contractual accountability. A provenance level is not a replacement for these broader vendor risk decisions. Organizations should document which questions provenance answers and which require other technical or business evidence.

Provenance can shorten incident scope when the evidence is usable

Suppose investigators discover that a hosted builder’s credentials were stolen. If the organization retains reliable provenance and stable artifact identities, it can identify which releases used the affected builder during the relevant time, which source revisions and inputs were involved, and where those artifacts were deployed. Teams can prioritize replacement of affected software rather than assuming every product released during the year is compromised.

That conclusion still needs qualification. An apparently unaffected artifact might have been built using a shared dependency or a different route controlled by the same adversary. A forged attestation or compromised signing authority can undermine evidence. Incident responders should check independently gathered build-platform logs, repository history, package distribution records and actual deployments. Provenance narrows questions when it is credible; it is not a substitute for investigation.

Plan for key and builder compromise before a crisis. Affected signer identities may need revocation, attestations may need to be distrusted or re-evaluated, and releases may need to be rebuilt on a known-good platform. Incident playbooks should connect the ownership of signing keys, build infrastructure, distribution services and production admission controls. Waiting until an emergency to determine who can withdraw trust from a compromised builder wastes critical time.

After remediation, verify that the new artifact has a new, approved digest, that its provenance identifies the replacement builder and source, and that production policy admits it while still rejecting the affected old release. A successful recompile does not prove remediation if the vulnerable build path or key remains in the pipeline.

Adopt track-specific evidence, then test it against realistic attacks

Begin by mapping how important applications move from source change to deployed artifact. Identify the owners of repositories, hosted build services, signing identities, artifact registries and deployment policy. Determine which records already exist and where identities or digests cannot be connected. The first improvement may be reliable Build L1 provenance and a consistent build process, followed by authenticated hosted provenance and hardened platform requirements as the organization gains the capability to enforce them.

In parallel, mature the Source Track by preserving trusted history, issuing appropriate source provenance, enforcing organizational controls and adding review according to the approved requirements. Record claims separately for Build and Source; do not turn the level names into an arbitrary project score. Verify at least one end-to-end release from source approval through production and include negative cases that should not pass the policy.

Metrics should express actual coverage and reliability: proportion of production artifacts bound to verified digests, unexpected builder identities detected, missing attestations, rejected untrusted releases, time to identify affected outputs after builder compromise, and exceptions still open. More provenance files are not inherently better if consumers never validate them. A program succeeds when the evidence changes what can safely enter production and supports faster investigations.

SLSA’s Build and Source Tracks describe evidence about how software moves from an authorized revision to a built artifact; their specific versioned levels are not separately listed as Security+ SY0-701 objectives. An organization should still distinguish source authorization, build isolation, authenticated provenance, component inventory and deployment policy. A valid attestation is useful only if consumers bind it to the exact artifact and enforce an independently chosen trust policy. Do not interpret a level as proof that application code is free of vulnerabilities.

Review question. A production service has a valid software signature and a complete SBOM, but the artifact digest differs from the digest in trusted build provenance. Can deployment proceed because two checks passed? No. The evidence does not describe the same output. Investigate the mismatch before accepting the artifact.

Leave a Reply

How It Works

img
Step 1. Choose Exam
on ExamLabs
Download IT Exams Questions & Answers
img
Step 2. Open Exam with
Avanset Exam Simulator
Press here to download VCE Exam Simulator that simulates real exam environment
img
Step 3. Study
& Pass
IT Exams Anywhere, Anytime!