feat: add Amp (ampcode.com) as a supported platform#790
Closed
zuwasi wants to merge 392 commits into
Closed
Conversation
…js, macOS watch, god_nodes degree rename Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…correct rustworkx APIs, git fetch strategy
…aphify-Labs#401 Graphify-Labs#385, team workflow docs, Windows/pipx tips Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…Graphify-Labs#418 graph.html missing from CLI
…raphify-Labs#423 --update drops existing nodes
…add .mdx to DOC_EXTENSIONS
…#260 add .html to DOC_EXTENSIONS
…s#431 Go import node ID collision
…h and save_semantic_cache, fix Graphify-Labs#454 sanitize_label crash on None source_file
…Graphify-Labs#436 _is_sensitive false positives on directory names, fix Graphify-Labs#433 update command writes absolute paths in artifacts Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…s in README Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…t/new_expression extraction
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…-035/F-038/PR747-NEW-2
…sing (Graphify-Labs#766) tree-sitter-typescript ships two grammars: - language_typescript: pure TypeScript, no JSX support - language_tsx: JSX-aware variant for .tsx files Currently both .ts and .tsx are parsed with language_typescript, which treats JSX syntax as parse errors. Every function declaration, arrow function, and call_expression nested inside a JSX tree is silently dropped from the extracted graph. Repro on a representative React+TypeScript codebase (a 13-file Tauri app): parsing each .tsx with language_typescript produces ~276 ERROR nodes per file. Only declarations that happen to live before the first JSX block survive. Fix: add _TSX_CONFIG that mirrors _TS_CONFIG but selects language_tsx, and route .tsx files to it in extract_js(). Effect on the same repo (graphify update --force): Nodes: 303 → 618 (+104%) Edges: 482 → 779 (+62%) Communities: 28 → 45 (+61%) Parse errors 276 → 0 per .tsx file Tests added: - tsx fixture with helpers + JSX-returning component - helpers and component are captured - JSX expression calls ({fmtDate(now)}) resolve to call edges - wiring check: .tsx uses language_tsx, .ts uses language_typescript Note: this fixes the parsing layer. Calls inside deeply nested arrow function callbacks (e.g. items.map(x => <T>{f(x)}</T>)) are still missed by the call extraction logic — separate enhancement. Co-authored-by: Serkan Gezici <serkan@quadroaipilot.com>
…oriented (Graphify-Labs#774) 10 skill-*.md files had descriptions that only described what graphify does (input->pipeline->output), not when agents should use it. This meant skills never loaded proactively on codebase questions. Changed to hybrid descriptions that retain the pipeline summary but add trigger conditions: 'Use when user asks any question about a codebase, project content, architecture, or file relationships'.
…ful graph-first directives (Graphify-Labs#775) Rewrote _CLAUDE_MD_SECTION, _AGENTS_MD_SECTION, and _GEMINI_MD_SECTION to use forceful ALWAYS/NEVER directives instead of soft suggestions. Agents must now consult the knowledge graph before file operations. Context: - Updated AGENTS.md template injected via _agents_install() - Updated CLAUDE.md template injected via claude install - Updated GEMINI.md template injected via gemini_install()
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…upport for SQL (Graphify-Labs#779) The SQL parser (`extract_sql`) previously only extracted foreign key relationships defined inline within CREATE TABLE column definitions. FK constraints added via ALTER TABLE ... ADD CONSTRAINT ... FOREIGN KEY ... REFERENCES were silently ignored. Additionally, `_obj_name()` only read the first identifier child of object_reference nodes, so schema-qualified names like `Sales.Customer` were truncated to just `Sales`. Changes: - Add `alter_table` handler to `walk()` that extracts FK edges from ALTER TABLE ... ADD CONSTRAINT ... FOREIGN KEY ... REFERENCES - Fix `_obj_name()` to read the full object_reference text, preserving schema-qualified names (e.g. `Sales.Customer`) - Fix inline FK resolution in create_table and _walk_from_refs to use full object_reference text instead of first identifier only
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Graphify-Labs#760, Graphify-Labs#725) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…aphify-Labs#789) Adaptive retry only recovered from `finish_reason="length"` (output truncation). It did not handle the other shape of overflow: the API rejecting the prompt outright with a 400 because the input plus `max_completion_tokens` doesn't fit in the model's context window. This shows up immediately on local OpenAI-compatible servers (LM Studio, llama.cpp, vLLM) where the default context is small (4K-32K) and a 60K-token chunk packed for cloud Kimi/Claude blows past it. Without retry the whole chunk fails with no output, even though the two halves would each fit cleanly. Catch a heuristic set of context-overflow exception messages, classify them as the same kind of recoverable failure as `finish_reason="length"`, and split-recurse on the same path. Single- file overflow returns an empty fragment so the rest of the corpus keeps running. Unrelated errors (rate limit, auth, etc.) still propagate. Tested with qwen3.5-9b on LM Studio (32K ctx) against a 215-file corpus where chunks 4-12 of 12 previously failed; with this change the overflowing chunks self-heal by splitting in half.
… guards (Graphify-Labs#788) Three independent Windows compatibility fixes shipped together because they all surface during the same first /graphify run on Windows. graphify/benchmark.py print_benchmark() unconditionally printed U+2500 (box-drawing) and U+2192 (rightwards arrow), which UnicodeEncodeError'd on stdouts that can't encode them — most notably the legacy Windows console at cp1252. New _safe() helper falls back to ASCII when the active stdout encoding can't carry the glyph; _hr() uses it. Two regression tests cover both paths and prove print_benchmark survives a cp1252-strict stream. graphify/extract.py ProcessPoolExecutor on Windows uses spawn, so worker subprocesses re-import the calling __main__. When the caller is `python -c "..."` or a script without an `if __name__ == "__main__":` guard, the workers recursively spawn themselves and the pool dies. The user-visible failure was a 290-line traceback ending in BrokenProcessPool, hiding the actual cause. _extract_parallel now catches BrokenProcessPool, prints a one-line warning that names the __main__-guard idiom, and returns False so the public extract() routes to the existing _extract_sequential fallback. Two tests cover the parallel-returns-False contract and the sequential fallback wiring. graphify/skill-windows.md Every `python -c "..."` block (30 in total) is replaced with a Write+run+delete pattern using PowerShell's literal here-string @'...'@. The old form was a quote-escaping minefield: any double-quote inside the Python source had to be backslash-escaped for the shell, and PowerShell's parser ate them inconsistently — failing on f-strings like `f'AST: {len(result["nodes"])} nodes'`. The new form passes Python source to disk literally, so what the model writes is what Python sees. The AST step's script template now includes an explicit `if __name__ == "__main__":` guard so multi-core extraction works even before the runtime fallback above kicks in. All 31 resulting heredoc blocks parse cleanly under `ast.parse`. Co-authored-by: Nauman Hameed <Nauman.Hameed@enghouse.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add skill-amp.md with Task tool subagent dispatch, AGENTS.md integration, and discrete confidence rubric - Add 'amp' to _PLATFORM_CONFIG and CLI dispatch in __main__.py - Update README.md with Amp in platform tables and intro Amp-Thread-ID: https://ampcode.com/threads/T-019e0cb6-2b4c-7430-859f-b44a34d67050 Co-authored-by: Amp <amp@ampcode.com>
…tory) Amp-Thread-ID: https://ampcode.com/threads/T-019e0cb6-2b4c-7430-859f-b44a34d67050 Co-authored-by: Amp <amp@ampcode.com>
Collaborator
|
Thanks for adding Amp support — it's a platform we want to include. However this PR bundles 50,000+ lines of unrelated changes (skill.md rewrite, dedup.py, extract/build/cluster/serve/watch modifications, 22 translation READMEs, worked examples) alongside the Amp integration, which makes it impossible to review safely. We're closing this and asking for a scoped resubmission. The Amp-specific code also has two bugs:
Please resubmit a focused PR containing only:
That scoped PR will merge quickly. Thanks! |
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
19 participants
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.
What
Adds Amp (Sourcegraph's AI coding agent) as a supported platform for graphify.
Changes
New file:
graphify/skill-amp.md(1016 lines)Amp-specific skill file adapted from the Claude Code skill, with:
graphify-out/prefix for all temp filesModified:
graphify/__main__.pyampto_PLATFORM_CONFIGwithskill_dst: ~/.amp/skills/graphify/SKILL.mdampto the agents-install CLI dispatch branchModified:
README.mdInstall (after merge)
How Amp dispatches subagents
Amp uses the
Tasktool for parallel subagent dispatch — multiple Task calls in the same message run concurrently. Sub-agents write chunk JSON files to disk usingcreate_file, then results are merged.Testing
Tested on Windows 11 with Amp CLI. The skill follows the same pipeline as skill.md with Amp-specific subagent dispatch.