agent-graph-workflow-v7 - #743
Closed
robertmonka wants to merge 3 commits into
Closed
Conversation
robertmonka
force-pushed
the
agent-graph-workflow-v7
branch
2 times, most recently
from
May 7, 2026 09:43
3f6dc3b to
d12919b
Compare
robertmonka
force-pushed
the
agent-graph-workflow-v7
branch
9 times, most recently
from
May 9, 2026 23:29
ebb4c91 to
b14ee33
Compare
Extend func_node tracking to Swift, Kotlin, Go, C#, and PHP call handlers so the AST node is consistently available. This fixes cases where caller/callee relationships were missing node references needed for downstream edge generation. Normalize both file node IDs and entity stem-prefix IDs when remapping absolute paths to project-relative paths. Entity IDs use _file_stem(path) as their prefix, so remapping only whole-file IDs left cross-file edges unresolved (Graphify-Labs#550).
Rename `graphify install` to `graphify skill <platform>` to clarify it installs assistant skills rather than package binaries. Add vscode platform support alongside existing claude, copilot, and others. Load environment variables via `graphify.env.load_env()` on CLI startup to support `GRAPHIFY_OUT` overrides and `CLAUDE_CONFIG_DIR` for custom skill destinations. Split generic `skill.md` into platform-specific files (`skill-claude.md`, `skill-vscode.md`) so each assistant receives tailored instructions. Update output path references from `.graphify_semantic_new.json` to `graphify-out/.graphify_semantic_new.json` to match the configurable output directory. Refactor platform config to use helper functions for source and destination paths, reducing duplication and enabling environment-based overrides for Claude skills.
Refactor Codex integration from deprecated hooks.json to config.toml with features.hooks flag, adding migration path for legacy setups. Clean legacy hook entries during install to avoid duplicates. Add Kimi Code to supported platforms list with skill directory installation. Include tomli dependency for Python < 3.11 compatibility to support tomllib fallback in tests and runtime.
robertmonka
force-pushed
the
agent-graph-workflow-v7
branch
from
May 10, 2026 06:53
b14ee33 to
51e7cc8
Compare
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
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.
🧢 Changes
graphify skill <platform>installs or refreshes the assistant skill.graphify setup <platform>writes project-local assistant instructions and hooks.graphify install --platform <platform>and platform-firstgraphify <platform> installforms..envloading for CLI and direct LLM usage, including project/home.envdiscovery, WSL Windows-home fallback, and Windows user environment support for common graphify/API variables.MOONSHOT_API_KEYis set, semantic extraction can usegraphify.llm.extract_corpus_parallel(..., backend="kimi")directly instead of dispatching Codex subagents.graphify-out/so skill runs avoid root-level.graphify_*scratch files.UserPromptSubmitreminder and aPreToolUseguard that blocks raw search/read/list tools until the graph has been used in the session.skillandsetupcommand model.☕️ Reasoning
The previous install surface mixed two different responsibilities: installing reusable assistant skills and configuring the current project to use a graph. This PR separates those flows while preserving old commands as aliases, which makes onboarding clearer without breaking existing users.
It also makes the graphify assistant workflow more reliable for larger or repeated runs. Kimi can be used directly when a Moonshot key is available, temporary files stay inside
graphify-out/, Claude is nudged or guarded toward graph-first exploration, and.envloading makes API-key configuration work more consistently across local, Windows, and WSL environments.🧪 Testing
.envloading precedence, quoted/exported values, process-env protection, and WSL Windows-home discovery.