Struct PackBuilder 

pub struct PackBuilder { /* private fields */ }

Implementations

impl PackBuilder

pub fn new() -> Self

pub fn preset(self, p: Preset) -> Self

pub fn budget(self, b: Budget) -> Self

pub fn max_tokens(self, n: usize) -> Self

pub fn reserve_tokens(self, n: usize) -> Self

pub fn tokenizer(self, t: Tokenizer) -> Self

pub fn scrub(self, on: bool) -> Self

pub fn expand_mode(self, m: ExpandMode) -> Self

pub fn include_path(self, path: impl Into<String>) -> Self

pub fn exclude_path(self, path: impl Into<String>) -> Self

pub fn project_root(self, p: impl Into<PathBuf>) -> Self

pub fn stdin_prompt(self, prompt: impl Into<String>) -> Self

pub fn files_from(self, paths: Vec<PathBuf>) -> Self

Repo-relative paths whose full contents should be embedded in a dedicated “📂 Scoped Files” section. Designed for the cargo impact --context | cargo context --files-from - workflow: the upstream tool tells us which files matter, and we include them verbatim (subject to scrubbing and budget).

These paths also join the diff-changed set when the related-tests collector runs, so a test that references a scoped file by stem gets surfaced even when the scoped file isn’t in git diff.

pub fn impact_findings(self, findings: Vec<Finding>) -> Self

Findings pulled from a cargo-impact --format=json envelope.

Differs from Self::files_from in that each finding carries metadata (id, kind, confidence, severity, tier, evidence, suggested action) that can drive richer rendering:

• In the default aggregated mode, the list is rendered as a single “📂 Scoped Files” section with files ordered by the caller-supplied order (typically confidence-descending) and per-file headers that surface confidence/severity/tier.
• When Self::impact_per_finding is set, each finding becomes its own “📂 Impact: …” section — useful when an agent wants to iterate through findings one at a time.

Finding paths join the diff-changed set for related-tests linkage, mirroring files_from.

pub fn impact_per_finding(self, on: bool) -> Self

Emit one pack section per finding instead of one aggregated section. No-op when impact_findings is empty.

pub fn build(self) -> Result<Pack>

Assemble the pack.

Flow:

1. Collect candidate sections per preset (collectors that fail for expected reasons return None — a missing git repo or Cargo.toml never aborts the build).
2. Run the scrubber over each section’s content. Scrubbing may shrink or grow content, so we re-count tokens after it.
3. Reconcile with the token budget using the configured strategy. Dropped section names are surfaced on Pack.dropped.

Trait Implementations

impl Clone for PackBuilder

fn clone(&self) -> PackBuilder

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for PackBuilder

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for PackBuilder

fn default() -> Self

Returns the “default value” for a type.