[file-diet] File Diet: Refactor pkg/cli/experiments_command.go (1120 lines)

[github-actions]

Overview

The file pkg/cli/experiments_command.go has grown to 1120 lines, making it difficult to maintain and test. This task involves refactoring it into smaller, focused files with improved test coverage.

Current State

• File: pkg/cli/experiments_command.go
• Size: 1120 lines
• Test Coverage: pkg/cli/experiments_command_test.go is 539 lines (~48% test-to-source ratio)
• Complexity: Contains ~40 top-level functions spanning several distinct concerns: CLI command wiring, list/analyze orchestration, remote & local data fetching (git branches, GitHub API), JSONL state parsing, git-ref safety validation, and console rendering.

Full File Analysis

Detailed Breakdown

Functions and their approximate responsibilities:

• Cobra command wiring (lines 94–195): NewExperimentsCommand, NewExperimentsListSubcommand, NewExperimentsAnalyzeSubcommand
• Command orchestration (lines 195–330): RunExperimentsList, RunExperimentsAnalyze, computeExperimentAnalyses
• Metric eval loading/summarizing (lines 331–411): loadLocalMetricEvalResults, loadRemoteMetricEvalResults, summarizeMetricEvalResults
• Experiment config discovery (lines 412–612): loadLocalExperimentConfigs, loadRemoteExperimentConfigs, findRemoteWorkflowFilenameForExperiment, matchWorkflowFilenameByExperiment, findWorkflowFileForExperiment, workflowFileCandidates
• Experiment listing/details fetch (local & remote) (lines 613–762): fetchLocalExperiments, fetchRemoteExperiments, fetchLocalExperimentDetails, fetchRemoteExperimentDetails, experimentStateFilenames, readLocalExperimentState
• Git ref/path safety validation (lines 762–851): buildSafeGitShowObjectArg, isSafeExperimentStateRef, isHexObjectIDPrefix, isSafeGitRefName, isSafeGitTreePath
• Experiment state parsing & aggregation (lines 851–1023): readRemoteExperimentState, appendExperimentRun, parseExperimentStateJSONL, parseExperimentState, emptyExperimentState, experimentInfoFromState, experimentDetailsFromState, experimentTotalRuns, experimentLastRun, extractExperimentName, gitRefExists
• Rendering (lines 1024–1106): printExperimentDetails, formatAssignments
• Generic helper (lines 1106+): parsePagedJSONArray[T any]

Notable coupling/complexity:

• Local vs. remote variants of many functions (fetchLocal*/fetchRemote*, loadLocal*/loadRemote*) duplicate structural logic while differing only in the data source (local git state vs. remote GitHub API), a strong signal for extracting a shared interface/abstraction.
• The git-ref/path safety validators (isSafe*) are a self-contained security-sensitive module that would benefit from isolation and dedicated, exhaustive testing.
• JSONL/state parsing functions are independent of CLI/cobra concerns and could be moved to a state-only file.

Refactoring Strategy

Proposed File Splits

1. experiments_command.go (kept, slimmed down)

   • Functions: NewExperimentsCommand, NewExperimentsListSubcommand, NewExperimentsAnalyzeSubcommand, RunExperimentsList, RunExperimentsAnalyze, computeExperimentAnalyses
   • Responsibility: Cobra command definitions and top-level orchestration only
   • Estimated LOC: ~250

2. experiments_fetch.go

   • Functions: fetchLocalExperiments, fetchRemoteExperiments, fetchLocalExperimentDetails, fetchRemoteExperimentDetails, loadLocalExperimentConfigs, loadRemoteExperimentConfigs, findRemoteWorkflowFilenameForExperiment, matchWorkflowFilenameByExperiment, findWorkflowFileForExperiment, workflowFileCandidates, loadLocalMetricEvalResults, loadRemoteMetricEvalResults, summarizeMetricEvalResults
   • Responsibility: Discovering and loading experiment configs/metrics from local git state and remote GitHub API
   • Estimated LOC: ~420

3. experiments_state.go

   • Functions: experimentStateFilenames, readLocalExperimentState, readRemoteExperimentState, appendExperimentRun, parseExperimentStateJSONL, parseExperimentState, emptyExperimentState, experimentInfoFromState, experimentDetailsFromState, experimentTotalRuns, experimentLastRun, extractExperimentName, gitRefExists
   • Responsibility: Parsing and aggregating experiment state (JSONL records → structured results), independent of CLI wiring
   • Estimated LOC: ~270

4. experiments_git_safety.go

   • Functions: buildSafeGitShowObjectArg, isSafeExperimentStateRef, isHexObjectIDPrefix, isSafeGitRefName, isSafeGitTreePath
   • Responsibility: Security-sensitive validation of git refs/paths used when reading experiment state via git show
   • Estimated LOC: ~90

5. experiments_render.go

   • Functions: printExperimentDetails, formatAssignments
   • Responsibility: Console rendering of experiment details
   • Estimated LOC: ~90

Shared Utilities

• experiments_json_utils.go: Extract parsePagedJSONArray[T any] here since it's a generic helper usable by any paged-JSON-array consumer beyond experiments (or move to an existing generic utils file if one already fits, e.g. a shared json_utils.go).

Interface Abstractions

• Introduce a small experimentDataSource interface (e.g., FetchExperiments() ([]ExperimentInfo, error), FetchExperimentDetails(id string) (*ExperimentDetails, error)) with localExperimentDataSource and remoteExperimentDataSource implementations. This removes the pervasive Local/Remote function-name duplication and lets RunExperimentsList/RunExperimentsAnalyze select an implementation based on repoOverride instead of branching per call site.

Test Coverage Plan

1. experiments_fetch_test.go

   • Test cases: local config discovery with matching/mismatched workflow IDs, remote fetch fallback paths, metric eval summarization with malformed/empty JSON, workflow filename matching heuristics
   • Target coverage: >80%

2. experiments_state_test.go

   • Test cases: JSONL parsing with valid/partial/corrupt lines, state aggregation across multiple runs, empty-state defaults, extractExperimentName edge cases (missing prefix, unusual ref formats)
   • Target coverage: >80%

3. experiments_git_safety_test.go

   • Test cases: valid/invalid git ref names, path traversal attempts in tree paths, hex object ID prefix boundary cases, injection attempts in buildSafeGitShowObjectArg
   • Target coverage: >90% (security-sensitive)

4. experiments_render_test.go

   • Test cases: formatting assignments with empty/singleton/multiple entries, printing details with nil/partial fields
   • Target coverage: >80%

Implementation Guidelines

1. Preserve Behavior: Ensure all existing functionality works identically
2. Maintain Exports: Keep public API unchanged (exported functions/types)
3. Add Tests First: Write tests for each new file before refactoring
4. Incremental Changes: Split one module at a time
5. Run Tests Frequently: Verify make test-unit passes after each split
6. Update Imports: Ensure all import paths are correct
7. Document Changes: Add comments explaining module boundaries

Acceptance Criteria

• Original file is split into 5 focused files
• Each new file is under 500 lines
• All tests pass (make test-unit)
• Test coverage is ≥80% for new files
• No breaking changes to public API
• Code passes linting (make lint)
• Build succeeds (make build)

Additional Context

• Repository Guidelines: Follow patterns in .github/agents/developer.instructions.agent.md
• Code Organization: Prefer many small files grouped by functionality
• Testing: Match existing test patterns in pkg/cli/*_test.go

---

Priority: Medium
Effort: Medium (5 new files, existing test coverage at ~48% needs to grow alongside the split)
Expected Impact: Improved maintainability, easier testing, reduced complexity

Generated by 🧹 Daily File Diet · auto · 89.6 AIC · ⌖ 11.3 AIC · ⊞ 10.1K · ◷

• expires on Aug 20, 2026, 5:06 AM UTC-08:00

[github-actions]

Caution

agentic threat detected
Threat detection flagged this output in warn mode. Manual review is REQUIRED before any follow-up automation.

Details

Potential security threats were detected in the agent output.

Review the workflow run logs for details.

🍪 Issue Monster selected this for Copilot

I've identified this issue as a good candidate for automated resolution and requested assignment to the Copilot coding agent.

If assignment succeeds, the Copilot coding agent will analyze the issue and create a pull request with the fix.

Om nom nom! 🍪

🍪 Om nom nom by Issue Monster · gpt54 · 2.05 AIC · ⌖ 11.5 AIC · ⊞ 8.8K · ◷

[github-actions]

🍪 Issue Monster selected this for Copilot

I identified this issue as a good candidate for automated resolution and requested assignment to the Copilot coding agent.

If assignment succeeds, the Copilot coding agent will analyze the issue and create a pull request with the fix.

Om nom nom! 🍪

🍪 Om nom nom by Issue Monster · gpt54 · 3.21 AIC · ⌖ 9.54 AIC · ⊞ 11.2K · ◷