chore: remove stale generated output artifacts and ignore them - #9
Merged
Conversation
Delete leftover tool output that predates the GRAPH.* output rename (#7/#8): graphs.md, graphs.json, and the untracked graph.html, plus the committed .coverage sqlite database. None of these are inputs to the build, tests, or docs — they were graphLM's own output committed by accident. Add .coverage, coverage.xml, and the GRAPH.* output files to .gitignore so generated artifacts no longer drift back into version control. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01SLftfcgzQfsWHbuA4XHv1F
ggrace519
added a commit
that referenced
this pull request
Aug 30, 2026
Adds the project CLAUDE.md (the /init deliverable), fact-checked by a multi-model review panel and corrected against the current code: - Commands: test count ~266; coverage is reported, not gated (no --cov-fail-under); mypy is a separate CI job, now in the dev group. - Two-pass / dry-run: dry-run selects file_fragments[:max_pass2_files] (cap 80) and reports graph-section counts, not "all files"; the pass-2 token budget is now enforced (edge table + instruction block reserved up front), so the description says so. - AST: project_dir is optional (resolution is against known_files, not the on-disk read); corrected the wrong causal claim. - Cycles: the risk term is a physical line count over the truncated fragment, not true SLOC. - HTML: single self-contained file but loads D3 from a CDN — not offline. - Security: sensitive-file rule now matches the fixed code (any dotenv except templates); symlink guard covers files as well as directories. - Config: documents the --max-context > GRAPHLM_MAX_CONTEXT > 120000 precedence; adds a CLI-flags section (incl. --no-show-cycles / --cycle-threshold). - Removed the now-false housekeeping note (those files were deleted in #9). Every changed claim verified against the code (flags exist, CDN URL, _ENV_SAFE_SUFFIXES, duplicate detect_import_cycles, 266 tests, --help). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01SLftfcgzQfsWHbuA4XHv1F
ggrace519
added a commit
that referenced
this pull request
Aug 30, 2026
#15) Adds the project CLAUDE.md (the /init deliverable), fact-checked by a multi-model review panel and corrected against the current code: - Commands: test count ~266; coverage is reported, not gated (no --cov-fail-under); mypy is a separate CI job, now in the dev group. - Two-pass / dry-run: dry-run selects file_fragments[:max_pass2_files] (cap 80) and reports graph-section counts, not "all files"; the pass-2 token budget is now enforced (edge table + instruction block reserved up front), so the description says so. - AST: project_dir is optional (resolution is against known_files, not the on-disk read); corrected the wrong causal claim. - Cycles: the risk term is a physical line count over the truncated fragment, not true SLOC. - HTML: single self-contained file but loads D3 from a CDN — not offline. - Security: sensitive-file rule now matches the fixed code (any dotenv except templates); symlink guard covers files as well as directories. - Config: documents the --max-context > GRAPHLM_MAX_CONTEXT > 120000 precedence; adds a CLI-flags section (incl. --no-show-cycles / --cycle-threshold). - Removed the now-false housekeeping note (those files were deleted in #9). Every changed claim verified against the code (flags exist, CDN URL, _ENV_SAFE_SUFFIXES, duplicate detect_import_cycles, 266 tests, --help). Claude-Session: https://claude.ai/code/session_01SLftfcgzQfsWHbuA4XHv1F Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Removes leftover graphLM output that was committed by accident and predates the
GRAPH.*output rename (#7/#8). None of these files are inputs to the build, tests, or docs — they are the tool's own generated output.What changed
graphs.md,graphs.json(tracked, old-named Markdown/JSON output), the untrackedgraph.html, and the committed.coveragesqlite database..gitignore: added.coverage,coverage.xml, and the currentGRAPH.md/GRAPH.json/GRAPH.htmloutput names so generated artifacts stay out of version control under both the old and new naming.Confirmed each deleted file was self-referential generated output (the
graphs.jsondirectory tree even listed itself and its siblings as scanned files). The only remaining mention ofgraph.htmlanywhere is a historical CHANGELOG line describing a past bug fix — correct as-is, left untouched.Verification
uv run pytest -q→ 261 passed (unchanged before and after removal)grepfor references found only self-references inside the deleted files plus the historical CHANGELOG linePost-merge note
CLAUDE.mdwas created in the same working session (the/initdeliverable) but is intentionally not in this PR — it is a separate docs theme and will go in its own PR.🤖 Generated with Claude Code