buildAnthropicParams converts the shared ChatRequest into Anthropic SDK parameters.

cancelPendingCompression aborts the conversation's in-flight background compression, if any.

dispatchSubtasks groups items into batches per the configured strategy, then processes batches sequentially while running files within each batch conc

ensureModelInList appends model to the end when missing; never reorders existing entries.

executeSubtask performs the Plan Phase + Main Loop for a single file. It returns (completed, stop, err): a hard Go error (err) for provider/config/ctx

extractFilenameFromURL returns the filename part (last path segment) of a URL pattern.

gitOut runs a git command in dir and returns its trimmed stdout, failing the test on error. Used to capture the SHAs a ResolveInput result must match.

hashFields folds a sequence of string fields into one SHA-256 digest, writing an unambiguous 8-byte big-endian length prefix before each field so no t

manifestItemID derives the content-independent item_id for a diff, keyed on operation, input mode and the normalized old/new paths. RegisterSelected a

mergeBackground combines the inline --background value (or an auto-populated commit message) with the content read from --background-file, separated b

planEnabled / dedupEnabled / summaryEnabled report whether each optional phase will actually run: template must define it AND the corresponding --no-*

Drive post() over a sequence of per-call fake_urlopen outcomes. Each outcome is either the raw response body (bytes, returned) or an

readRuleFileSafe reads and validates a rule file. It enforces extension whitelist (.md / .txt / .markdown), a 512 KB size cap, and resolves symlinks b

AwaitKey blocks until every unit submitted under key so far has completed. It never touches the pool-wide WaitGroup, so it is safe to call while other

DefaultConfig returns a disabled default configuration.

HasInclude reports whether any include patterns are configured.

IsReserved reports whether name matches any built-in tool name (including Unknown).

LoadAppConfig loads config from path. Returns nil, nil if file does not exist.

LoadGitignorePatterns reads and parses .gitignore patterns from the given repository root. Returns nil if the file is missing or unreadable.

LoadSummary loads a single session's Summary. Errors when the session file is missing or unreadable.

ParseExtraHeaders parses a string of comma-separated key=value pairs into a dictionary. Values may be double-quoted to include commas, e.g. X-Forwarde

PrintTraceSummary prints a one-line summary of the review to stdout.

ProjectSummary is the markdown project-level summary produced by scan's PROJECT_SUMMARY_TASK. Empty for review mode and for scans that skipped / faile

RecordToolResult sets the outcome of a tool execution on the span.

RecordWarning adds a non-fatal warning.

SetFinalManifest stores the frozen manifest the agent produced. It is embedded into session_end by Finalize and returned to the CLI via FinalManifest.

Shutdown flushes and shuts down all initialized providers. It should be called before process exit to ensure buffered data is exported.

Submit runs f in a background goroutine bounded by the semaphore. When f completes its return value is collected internally.

ToToolDef converts an MCP tool definition to an llm.ToolDef.

TotalInputTokens returns the accumulated input/prompt tokens from all LLM calls.

UseTestSessions redirects session persistence to the "test-sessions" subdirectory so that test runs do not pollute the real sessions store. It must b

WithinBase reports whether target is base itself or contained under base.

buildAllowedHosts returns the default-deny allowlist used by the host guard. The set always includes the standard loopback names. If bindHost is a con

emitFailureUsage writes a best-effort structured usage record to stderr when a review fails, so the outer caller still sees the cost of the failed att

ensureMessagesSuffix appends /v1/messages to base URLs that lack a versioned path.

estimateDiffFileTokens projects the input+output token cost of reviewing a single diff (PLAN + MAIN_TASK rounds). It mirrors scan.estimateFileTokens b

executeSubtask runs the scan pipeline for one file: 1. Optional PLAN_TASK: produce a JSON checklist of focus areas. 2. MAIN_TASK: review the file with

filterLargeDiffs drops diffs whose diff content alone consumes more than 80% of MaxTokens.

filterLargeScans drops items whose content exceeds 80% of MaxTokens.

finalizeManifest freezes the run's coverage builder and stores the immutable manifest on the session for persistence and CLI consumption. Nil-safe. A

formatPlanGuidance parses the PLAN_TASK JSON output into a markdown snippet suitable for embedding in MAIN_TASK. On parse failure it returns the raw c

humanTokens formats a token count as e.g. "1.2M" / "850K" / "420".

humanTokens formats a token count as e.g. "1.2M" / "850K" / "420". Mirrors internal/scan/estimate.go; kept here so agent has no scan import.

maybeRunDedup, when the template has a DedupTask and the batch produced at least DedupMinComments comments, invokes the DEDUP_TASK LLM to merge near-d

maybeRunProjectSummary runs the post-batch PROJECT_SUMMARY_TASK over the union of all collected comments. Best-effort: any error / empty input no-temp

normalizeLine removes leading/trailing whitespace and strips any leading '+' or '-' diff marker.