Aura Code Audit Trail

Track what coding agents changed and why, then block commits whose symbol-level changes violate their stated intent.

Source repo
Naridon-Inc/aura
Stars
★ 47
Last updated
5d ago
License
Apache-2.0
Primary language
Rust

At a glance

Works with
Universal · cross-platformCodex · Claude Code
You'll need
GitRust and Cargo when building from sourceShell / CLILocal filesystemMCP Server
Typical use
A team running several coding agents needs to determine months later which agent changed a function, what request initiated the work, and what intent was recorded.
Main limitation
Adoption adds .aura/ data, Git notes, and commit hooks to the repository workflow, all of which the team must account for operationally.
Source review
70/100 · Some gaps

What does this agent do, and when should you use it?

Aura adds a semantic audit and versioning layer on top of Git, recording which symbols an agent or developer changed, the stated reason, and signed provenance for each commit. A single Rust engine powers the `aura` CLI, a Tauri and React desktop application, and a VS Code extension, all of which share the same `.aura/` meaning plane. Git hooks parse each commit with Tree-sitter, represent functions and classes in an AST Merkle graph with call edges, and compare the resulting changes with the declared intent. Its outputs include queryable provenance, semantic-review findings, signed metadata, AI usage and cost records, and function-level rewinds that do not revert the surrounding commit. Data stays in `.aura/` and Git notes by default, requires no server, and leaves the machine only when cloud sync is configured.

After aura init installs Git hooks, Aura parses every commit into logic nodes, computes AST-level differences, and links those differences to the intent recorded while the commit is being written. Its pre-commit gatekeeper can reject a commit when the stated task and the affected code disagree, such as an intent concerning login that modifies billing. aura why <file> reports the landed commit, author or agent, originating request, recorded intent, and capture method for a file or line. The CLI also exposes semantic review, AST-node-based function rewind, the Crew work loop, AI usage and cost tracking, and an MCP server started with aura mcp; the desktop and VS Code interfaces use the same .aura/ records. Tree-sitter analysis is documented for Rust, Python, TypeScript, JavaScript, Go, Java, C#, C++, C, Ruby, PHP, Swift, and Kotlin.

  1. A team running several coding agents needs to determine months later which agent changed a function, what request initiated the work, and what intent was recorded.
  2. An engineering organization with audit obligations needs signed, symbol-level provenance and a stated reason for code changes.
  3. A maintainer wants to stop out-of-scope edits before commit, such as an agent deleting authentication logic while claiming to tune retry behavior.
  4. A developer needs to restore one faulty function without reverting unrelated work contained in the same commit.
  5. A large-codebase review team wants to detect layer violations, silent deletions, and architectural drift from graph and AST changes rather than text alone.
  6. A team using Claude Code, Gemini CLI, Codex CLI, Kimi, Cursor, or OpenCode wants repository-local records covering changes produced by several tools.

How do you install or deploy this agent?

The prebuilt installer is documented for macOS, Linux, and Windows and requires a shell capable of running the Bash command:

curl -fsSL https://auravcs.com/install.sh | bash
cd your-repo && aura init

aura init installs hooks in the current Git repository. To build from source, install Rust and Cargo, then run this from the repository root:

cargo build --release

No server or cloud credentials are required for the local workflow. The supplied material does not document the account, credential, or configuration requirements for optional cloud sync.

How do you use this agent?

Once initialized, create commits normally; the hooks perform AST parsing, semantic diffing, intent logging, and gate checks. Query the reason behind changes to a file with:

aura why src/billing.rs

For direct access from an MCP-capable coding agent, add this server definition to the client's MCP configuration:

{
  "mcpServers": {
    "aura-vcs": { "command": "aura", "args": ["mcp"] }
  }
}

The MCP server exposes more than 30 tools, including aura_usage for agent spend monitoring. Telemetry can be disabled with either documented environment variable:

export AURA_TELEMETRY_OPTOUT=1
export DO_NOT_TRACK=1

What are this agent's strengths and limitations?

Pros
  • It records intent, provenance, and signatures at function and class level instead of limiting history to changed text lines.
  • The pre-commit gatekeeper compares declared intent with actual AST changes and can stop task drift before a commit lands.
  • Function-level rewind operates on AST nodes, allowing one function to be restored without reverting neighboring work or creating text merge conflicts for that operation.
  • The CLI, desktop application, and VS Code extension share one .aura/ meaning plane, so records remain consistent across interfaces.
  • Core records live in .aura/ and Git notes, allowing local operation without deploying a server.
Limitations
  • Adoption adds .aura/ data, Git notes, and commit hooks to the repository workflow, all of which the team must account for operationally.
  • Semantic parsing is documented for 13 languages; the source provides no evidence that other languages receive equivalent AST-level auditing or rewind support.
  • Integration depth varies by client: Claude Code has MCP tools, a status line, and transcript parsing, while Cursor is only explicitly associated with workspace detection.
  • Cloud sync is mentioned but its deployment, credentials, pricing, and failure handling are not documented in the supplied material.
  • Source builds require Rust and Cargo, while the desktop application and VS Code extension have their own component-specific build locations.

How does this agent compare with similar options?

Aura complements rather than replaces Git. Git continues to manage bytes, text diffs, commits, branches, history, and remotes; Aura adds an AST Merkle graph, logic nodes, call edges, intent, provenance, and signed metadata. Compared with the line ownership shown by git blame, Aura is designed to identify the developer or agent responsible for a symbol-level change, explain why it was made, and check whether the implementation matched its declared intent.

Key facts side by side with the most closely related agents.

Agent Source review Stars Updated Language Full support on
Aura Code Audit Trail This agent 70 · Some gaps ★ 47 5d ago Rust Codex · Claude Code
Argot Repository Style Analyzer 93 · Excellent ★ 48 14d ago Rust Claude Code
git-lrc Commit-Time Code Review 80 · Good ★ 1.5k 7d ago Go Claude Code · OpenAI API · Claude API
Gortex Code Intelligence 78 · Good ★ 1.6k 7d ago Go Codex · Claude Code · OpenAI API · Claude API

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Some gaps
70/ 100 5-point scale 3.5 / 5
Trust 16/29
Reliability 9/14
Adaptability 14/18
Convention 14/18
Effectiveness 12/13
Verifiability 5/8
Why each dimension lost points
Trust16 / 29 · 2.8/5

The README identifies local storage, Git hooks, Git notes, optional cloud sync, telemetry opt-out, intent gating, signed attribution, and function-level rewind. Desktop tests require failed termination to preserve a usable session handle and offer retry. Deductions apply because the concrete permission boundaries and per-action confirmations for 30-plus MCP tools, agent processes, transcript access, networking, and cloud sync are not shown. The claim that nothing leaves the machine without cloud sync is also insufficiently reconciled with opt-out telemetry. Sensitive-data protection is mainly asserted through summaries and dependency comments, without redaction, key-storage, or encryption implementation. Dependencies are versioned and internal crates are vendored, but no lockfile, vulnerability audit, or automated dependency-update evidence is supplied.

Reliability9 / 14 · 3.2/5

The manifest explains vendoring internal crates so fresh clones can resolve, while CI builds and runs end-to-end and unit tests on Linux and macOS. Desktop tests specify strong failure behavior for unsuccessful process termination, unreachable daemons, unsupported CLI commands, and transcript fallback, earning full credit for failure messages. Deductions apply because Windows is claimed in the README but absent from both the supplied release and end-to-end matrices. The separately built stackgraph binary is excluded from the workspace and invoked as a subprocess, with incomplete evidence of how its availability is ensured. No tests were executed in this review.

Adaptability14 / 18 · 3.9/5

The documentation addresses auditing, human and agent commits, semantic review, rewind, desktop, CLI, VS Code, MCP, several agent products, and 13 programming languages. Tests also demonstrate agent-specific commands and conservative fallbacks. Deductions apply because exact activation rules and false-positive handling for agent detection, commit blocking, cloud synchronization, and semantic analysis are not fully documented, and the Windows environment claim lacks corresponding CI and release evidence.

Convention14 / 18 · 3.9/5

The README has strong organization, with quick start, architecture, repository layout, FAQ, privacy, integrations, and build guidance. Package version 0.19.52, tag-triggered generated release notes, and a complete Apache-2.0 license provide solid conventional metadata. Deductions apply because the curl-pipe installer lacks checksum, pinned-version, and uninstall guidance, while desktop and extension installation details are not present. Known limitations are scattered through the security policy and test commentary instead of being summarized for users. The maintainer organization, contribution route, and private vulnerability-reporting path are named, but publisher identity remains unverified by the stated registry and the fallback security contact is not concrete.

Effectiveness12 / 13 · 4.6/5

The example output connects commits, requests, intent, agent identity, and time. Commit gating, symbol-level audit, semantic review, and function-level rewind offer material value beyond ordinary textual Git history. Detailed desktop tests support readable choices, cross-agent chat, recovery controls, and layout usability. Deductions apply because there is no quantified evidence for performance, repository growth, parsing overhead, false-positive rates, telemetry cost, or cloud-sync cost, and the operational burden of the broad feature set is not discussed.

Verifiability5 / 8 · 3.1/5

Major claims are tied to named commands, directories, configuration locations, Cargo dependencies, CI jobs, and behavior-focused tests. Some tests explicitly distinguish runtime-unknown information from statically supported facts. Deductions apply because central security and effectiveness claims still rely heavily on README statements and comments; the supplied evidence omits core implementations, the lockfile, signature-verification tests, redaction tests, and release artifact verification. Absolute claims such as conflict-free rewind, auditor-ready provenance, and no data leaving the machine are therefore not fully corroborated.

Risks and how to mitigate them
  • The installer pipes a remote script directly into a shell; pin a version and independently inspect the script, checksums, and release signatures before adoption.
  • The local-only privacy claim coexists with opt-out telemetry; confirm the default fields, endpoints, triggers, and retention period.
  • Aura installs Git hooks, reads agent transcripts, runs agent processes, and exposes many MCP tools; review its command and filesystem boundaries in an isolated repository.
  • Windows support is not demonstrated by the supplied CI or release matrices, and this pre-1.0 project supports only its latest release line.
  • Conflict-free symbol rewind, signed provenance, and commit-blocking effectiveness were not independently verified by the supplied implementation files or an executed run.
Evidence confidence: Low Reviewed Sep 24, 2026 Reviewed revision 3708210dbe70
See the full review method →

FAQ

Does Aura send code or audit records to an external service?
Not by default. Records stay in local .aura/ storage and Git notes, and data leaves the machine only if cloud sync is configured. Usage tracking reads local Claude Code transcripts and makes no API calls.
Will it replace or rewrite the existing Git workflow?
No. Existing history, remotes, and workflows remain in place. Aura adds semantic records through Git hooks, .aura/, and Git notes.
Is Claude Code required?
No. Aura explicitly detects Claude Code, Gemini CLI, Codex CLI, Kimi, Cursor, and OpenCode, and its Git-hook capture path works with anything that commits. The depth of native integration differs between tools.
Do intent checks and rewind work for human-written commits?
Yes. The intent gatekeeper, function-level rewind, and semantic review also operate on human commits; provenance identifies the developer rather than an agent.
Does local use require a cloud account or persistent server?
No. Both versioning planes live with the repository and the core local workflow requires no server. Cloud sync is optional.
View on GitHub ↗ Install ↓

Compare agents like this one

The same FARS review applied across the shortlist this agent qualifies for.

Related agents