Feature/jcode integration by akvarel · Pull Request #2578 · Graphify-Labs/graphify

akvarel · 2026-08-09T19:38:29Z

No description provided.

graphify-labs

Graphify reviewed this change.

Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.

Graphify review — findings

This PR adds support for a new "Jcode" platform in the graphify install/uninstall tooling. It registers a Jcode Agent Skill, wires up a jcode-hook pre-tool guard (invoked via a new CLI command) that redirects raw search/read tool calls toward graphify query, and adds config.toml manipulation helpers to insert/remove the hook from Jcode's hooks.pre_tool setting. It also updates the README, CLI usage text, __main__ imports, uninstall-all cleanup, and a .gitattributes merge driver entry for graph.json. The surface area spans graphify/install.py (platform config, hook install/uninstall, TOML editing helpers), graphify/cli.py (hook guard logic and command dispatch), graphify/__main__.py (imports and usage), plus documentation and gitattributes.

Worth a look

jcode-hook guard blocks 'read' tool without checking target is indexed — graphify/cli.py · Escalate · medium
- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 488 functions depend on the 237 functions this change touches.

Health — this change adds coupling hotspots:

worse: dispatch_command() — 2 callers, 118 callees
worse: dispatch_install_cli() — 2 callers, 32 callees
worse: install() — 5 callers, 12 callees
worse: uninstall_all() — 2 callers, 14 callees

Verification — 488 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

verification_scope: 488 function(s) in the blast radius were not formally verified this run

· 4 more finding(s) on lines outside this diff (see the check run).

graphify-labs

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

Formal verification. 1 change(s) alter behavior, breaking input(s) attached.

Behavior changes: \_node\_search\_text changes behavior, here is the input that shows it.

The verifier found a concrete input on which \_node\_search\_text behaves differently before and after the change. If that change is intended, ship it; if not, this is your bug.

Guarantee: This difference was REPRODUCED, the verifier actually ran both versions on that input and saw them disagree. It is real, not an artifact.

Evidence: On input \{"data":"\{'n': 0, 's': 'x', 'l': \[1, 2\]\}","nid":"'a longer string'"\}, the old code produced '\\x00\\x00a longer string\\x00\\x00' but the new code produces '\\x00\\x00a longer string\\x00\\x00\\x00'. Paste that input straight into a regression test.

Graphify review — findings

This pull request adds native ingestion of lat.md/ validated-knowledge Markdown into Graphify's graph, so that headings become knowledge-section nodes with summaries, wiki links, source links, and @lat code references are wired into the code graph. It introduces a new check-knowledge CLI command (with a --json option) for validating link/reference integrity, plus a supporting graphify/lattice_ingest.py module, new tests, and TDD documentation. Separately, it adds Jcode as a supported install platform, including install/uninstall hook functions, a Jcode hook guard in the CLI, and related help-text and README updates. It also adds a git merge attribute for graphify-out/graph.json.

No blocking issues surfaced.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 1999 functions depend on the 594 functions this change touches.

Health — this change adds coupling hotspots:

worse: extract() — 388 callers, 43 callees
worse: dispatch_command() — 2 callers, 119 callees
worse: _get_extractor() — 26 callers, 7 callees
worse: _query_graph_text() — 20 callers, 8 callees
worse: dispatch_install_cli() — 2 callers, 32 callees
worse: install() — 5 callers, 12 callees
new: validate_lattice() — 7 callers, 7 callees
worse: uninstall_all() — 2 callers, 14 callees
…and 2 more

Verification — 1999 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

verification_scope: 1999 function(s) in the blast radius were not formally verified this run

Formal verification

Behavior changes: \_node\_search\_text changes behavior, here is the input that shows it.

The verifier found a concrete input on which \_node\_search\_text behaves differently before and after the change. If that change is intended, ship it; if not, this is your bug.

Guarantee: This difference was REPRODUCED, the verifier actually ran both versions on that input and saw them disagree. It is real, not an artifact.

Evidence: On input \{"data":"\{'n': 0, 's': 'x', 'l': \[1, 2\]\}","nid":"'a longer string'"\}, the old code produced '\\x00\\x00a longer string\\x00\\x00' but the new code produces '\\x00\\x00a longer string\\x00\\x00\\x00'. Paste that input straight into a regression test.

No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).

The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

Could not verify: Could not verify dispatch\_command.

The verifier did not have enough to check dispatch\_command, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 9 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly SystemExit — names the real obstacle, not a sampling gap)

Could not verify: Could not verify \_get\_extractor.

The verifier did not have enough to check \_get\_extractor, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: parameter `path` is annotated `Path` — outside the synthesizable primitive/collection set

Could not verify: Could not verify extract.

The verifier did not have enough to check extract, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: parameter `cache_root` is annotated `Path | None` — outside the synthesizable primitive/collection set

No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).

The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

Could not verify: Could not verify install.

The verifier did not have enough to check install, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 9 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly SystemExit — names the real obstacle, not a sampling gap)

Could not verify: Could not verify uninstall\_all.

The verifier did not have enough to check uninstall\_all, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: the input domain has 28 values but only 14 distinct were tested — a small finite domain must be EXHAUSTED, not sampled (an untested input could invert the result)

No difference found (not proven): No behavior difference found in \_compute\_idf (not a proof).

The verifier ran both versions of \_compute\_idf on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

Could not verify: Could not verify \_score\_query.

The verifier did not have enough to check \_score\_query, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 35 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly TypeError — names the real obstacle, not a sampling gap)

No difference found (not proven): No behavior difference found in \_subgraph\_to\_text (not a proof).

The verifier ran both versions of \_subgraph\_to\_text on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

· 4 grounded finding(s) anchored inline below; 6 more finding(s) on lines outside this diff (see the check run).

graphify-labs · 2026-08-11T10:30:23Z

@@ -4719,6 +4726,11 @@ def _is_cpp_header(path: Path) -> bool:

 def _get_extractor(path: Path) -> Any | None:


⚠️ Health regression — _get_extractor()

fans out to 7 callees (efferent coupling); 26 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-11T10:30:23Z

+    return resolved, None
+
+
+def extract_lattice_markdown(path: Path) -> dict[str, Any]:


⚠️ Health regression — extract_lattice_markdown()

fans out to 7 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-11T10:30:23Z

+            edge["resolved_knowledge_target"] = resolved
+
+
+def extract_lattice_code_ref_edges(


⚠️ Health regression — extract_lattice_code_ref_edges()

high coupling complexity (Ca·Ce = 15).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-11T10:30:23Z

+    )
+
+
+def validate_lattice(project_root: Path) -> dict[str, Any]:


⚠️ Health regression — validate_lattice()

fans out to 7 callees (efferent coupling); 7 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs-staging

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

Graphify review — findings

This pull request adds native ingestion and validation of lat.md/ "validated knowledge" directories into Graphify's graph, so Markdown headings, summaries, wiki links, source references, and @lat code-mention comments become graph nodes/edges queryable alongside code symbols. It introduces a new lattice_ingest module, a check-knowledge CLI command (with --json), automatic validation during graphify update, and touches extraction, serving/query, and CLI/main wiring. It also adds a new "Jcode" install platform (_install_jcode_hook/_uninstall_jcode_hook), related README/docs updates, TDD evidence documentation, and a .gitattributes merge driver for graph.json.

No blocking issues surfaced. 1 lower-confidence candidate did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 2006 functions depend on the 601 functions this change touches.

Health — this change adds coupling hotspots:

worse: extract() — 388 callers, 43 callees
worse: dispatch_command() — 5 callers, 120 callees
worse: _get_extractor() — 26 callers, 7 callees
worse: _query_graph_text() — 20 callers, 8 callees
worse: dispatch_install_cli() — 2 callers, 32 callees
worse: install() — 5 callers, 12 callees
new: validate_lattice() — 8 callers, 7 callees
worse: uninstall_all() — 2 callers, 14 callees
…and 1 more

Verification — 2006 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

verification_scope: 2006 function(s) in the blast radius were not formally verified this run

· 4 grounded finding(s) anchored inline below; 5 more finding(s) on lines outside this diff (see the check run).

graphify-labs-staging · 2026-08-11T10:48:13Z

+    )
+
+
 def dispatch_command(cmd: str) -> None:


⚠️ Health regression — dispatch_command()

fans out to 120 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs-staging · 2026-08-11T10:48:13Z

@@ -4719,6 +4726,11 @@ def _is_cpp_header(path: Path) -> bool:

 def _get_extractor(path: Path) -> Any | None:


⚠️ Health regression — _get_extractor()

fans out to 7 callees (efferent coupling); 26 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs-staging · 2026-08-11T10:48:13Z

+    return resolved, None
+
+
+def extract_lattice_markdown(path: Path) -> dict[str, Any]:


⚠️ Health regression — extract_lattice_markdown()

fans out to 7 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs-staging · 2026-08-11T10:48:13Z

+    )
+
+
+def validate_lattice(project_root: Path) -> dict[str, Any]:


⚠️ Health regression — validate_lattice()

fans out to 7 callees (efferent coupling); 8 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

Graphify review — findings

This PR adds native lat.md/ "validated knowledge" ingestion to Graphify, so Markdown headings, summaries, wiki links, source references, and @lat code-mention comments are parsed into the code graph. It introduces a new graphify/lattice_ingest module plus supporting changes across extraction, serving/query scoring, and the CLI (including a new check-knowledge command and automatic lattice validation during update), along with a merge attribute for graph.json, README/TDD documentation, and tests. It also adds a new Jcode install/uninstall integration path (_install_jcode_hook/_uninstall_jcode_hook) wired into __main__ and the CLI help/platform listings, with accompanying test_jcode coverage. The surface area spans install, extract, serve, CLI, lattice ingestion, docs, and gitattributes.

No blocking issues surfaced. 1 lower-confidence candidate did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 2098 functions depend on the 606 functions this change touches.

Health — this change adds coupling hotspots:

worse: extract() — 440 callers, 45 callees
worse: dispatch_command() — 5 callers, 120 callees
worse: _get_extractor() — 26 callers, 7 callees
worse: _query_graph_text() — 20 callers, 8 callees
worse: dispatch_install_cli() — 2 callers, 32 callees
worse: install() — 5 callers, 12 callees
new: validate_lattice() — 8 callers, 7 callees
worse: uninstall_all() — 2 callers, 14 callees
…and 1 more

Verification — 2098 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

verification_scope: 2098 function(s) in the blast radius were not formally verified this run

· 4 grounded finding(s) anchored inline below; 5 more finding(s) on lines outside this diff (see the check run).

graphify-labs · 2026-08-12T13:12:16Z

+    )
+
+
 def dispatch_command(cmd: str) -> None:


⚠️ Health regression — dispatch_command()

fans out to 120 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-12T13:12:16Z

@@ -4958,6 +4965,11 @@ def _is_cpp_header(path: Path) -> bool:

 def _get_extractor(path: Path) -> Any | None:


⚠️ Health regression — _get_extractor()

fans out to 7 callees (efferent coupling); 26 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-12T13:12:16Z

+    return resolved, None
+
+
+def extract_lattice_markdown(path: Path) -> dict[str, Any]:


⚠️ Health regression — extract_lattice_markdown()

fans out to 7 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-08-12T13:12:17Z

+    )
+
+
+def validate_lattice(project_root: Path) -> dict[str, Any]:


⚠️ Health regression — validate_lattice()

fans out to 7 callees (efferent coupling); 8 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

Formal verification. 1 change(s) alter behavior, breaking input(s) attached.

Behavior changes: \_node\_search\_text changes behavior, here is the input that shows it.

The verifier found a concrete input on which \_node\_search\_text behaves differently before and after the change. If that change is intended, ship it; if not, this is your bug.

Guarantee: This difference was REPRODUCED, the verifier actually ran both versions on that input and saw them disagree. It is real, not an artifact.

Evidence: On input \{"data":"\{'n': 0, 's': 'x', 'l': \[1, 2\]\}","nid":"'a longer string'"\}, the old code produced '\\x00\\x00a longer string\\x00\\x00' but the new code produces '\\x00\\x00a longer string\\x00\\x00\\x00'. Paste that input straight into a regression test.

Graphify review — findings

This PR adds native support for ingesting "validated knowledge" from a project-level lat.md/ directory into the graph. It introduces a new lattice_ingest module along with a check-knowledge CLI command, wires lattice validation into the update workflow, and adds parsing of Markdown headings, wiki links, source references, and @lat code-mention comments into graph nodes and edges. The changes span extraction, serving/query, CLI dispatch, and install code, and also add a new Jcode install/uninstall platform, documentation (README + TDD evidence), a merge attribute for graph.json, and a corresponding test suite.

No blocking issues surfaced. 1 lower-confidence candidate did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 2111 functions depend on the 606 functions this change touches.

Health — this change adds coupling hotspots:

worse: extract() — 445 callers, 45 callees
worse: dispatch_command() — 5 callers, 120 callees
worse: _get_extractor() — 26 callers, 7 callees
worse: _query_graph_text() — 20 callers, 8 callees
worse: dispatch_install_cli() — 2 callers, 32 callees
worse: install() — 5 callers, 12 callees
new: validate_lattice() — 8 callers, 7 callees
worse: uninstall_all() — 2 callers, 14 callees
…and 1 more

Verification — 2111 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

verification_scope: 2111 function(s) in the blast radius were not formally verified this run

Formal verification

Behavior changes: \_node\_search\_text changes behavior, here is the input that shows it.

The verifier found a concrete input on which \_node\_search\_text behaves differently before and after the change. If that change is intended, ship it; if not, this is your bug.

Guarantee: This difference was REPRODUCED, the verifier actually ran both versions on that input and saw them disagree. It is real, not an artifact.

Evidence: On input \{"data":"\{'n': 0, 's': 'x', 'l': \[1, 2\]\}","nid":"'a longer string'"\}, the old code produced '\\x00\\x00a longer string\\x00\\x00' but the new code produces '\\x00\\x00a longer string\\x00\\x00\\x00'. Paste that input straight into a regression test.

Could not verify: Could not verify dispatch\_command.

The verifier did not have enough to check dispatch\_command, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 23 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly SystemExit — names the real obstacle, not a sampling gap)

Could not verify: Could not verify extract.

The verifier did not have enough to check extract, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: parameter `cache_root` is annotated `Path | None` — outside the synthesizable primitive/collection set

Could not verify: Could not verify \_get\_extractor.

The verifier did not have enough to check \_get\_extractor, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: parameter `path` is annotated `Path` — outside the synthesizable primitive/collection set

No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).

The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

Could not verify: Could not verify install.

The verifier did not have enough to check install, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 23 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly SystemExit — names the real obstacle, not a sampling gap)

No difference found (not proven): No behavior difference found in uninstall\_all (not a proof).

The verifier ran both versions of uninstall\_all on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.