Wire availableLibraries into taskitem generation dispatch

This commit is contained in:
Bill
2026-03-03 09:50:05 -07:00
parent 0592c8c3a6
commit e24f677dde
3 changed files with 107 additions and 11 deletions

View File

@@ -39,6 +39,9 @@
#include "ASTFeatureExtractor.h"
#include "LanguageIdiomProfile.h"
#include "LanguageFitnessScorer.h"
#include "TaskitemLibraryAnnotator.h"
#include "Sprint286IntegrationSummary.h"
#include "Sprint288IntegrationSummary.h"
#include "PolyglotProjectSpec.h"
#include "ABIBoundaryExtractor.h"
#include "FFIGlueDispatcher.h"

View File

@@ -18,6 +18,8 @@
{"description", "Normalized requirements from whetstone_architect_intake."}}},
{"conflicts", {{"type", "array"},
{"description", "Optional requirement conflicts from whetstone_architect_intake."}}},
{"availableLibraries", {{"type", "array"},
{"description", "Optional candidate libraries for deterministic library/API dispatch."}}},
{"strictExecutionContract", {{"type", "boolean"},
{"description", "Require concrete execution contract metadata (stepIds/files/tools/tests)."}}}
}}, {"required", json::array({"normalizedRequirements"})}}
@@ -477,10 +479,50 @@
};
}
std::vector<std::string> availableLibraries;
if (args.contains("availableLibraries")) {
if (!args["availableLibraries"].is_array()) {
return {
{"success", false},
{"error", "available_libraries_not_array"}
};
}
for (const auto& libJson : args["availableLibraries"]) {
if (!libJson.is_string()) {
return {
{"success", false},
{"error", "available_library_entry_not_string"}
};
}
const std::string lib = libJson.get<std::string>();
if (lib.empty()) continue;
bool seen = false;
for (const auto& existing : availableLibraries) {
if (existing == lib) {
seen = true;
break;
}
}
if (!seen) availableLibraries.push_back(lib);
}
}
const bool libraryDispatchEnabled = !availableLibraries.empty();
auto capabilityLedger = whetstone::Sprint286IntegrationSummary::buildSeedLedger();
auto symbolCatalog = whetstone::Sprint288IntegrationSummary::buildSeedCatalog();
whetstone::TaskitemLibraryAnnotator libraryAnnotator;
std::string libraryDispatchContext;
for (const auto& requirement : normalized.requirements) {
if (!requirement.normalizedText.empty()) {
if (!libraryDispatchContext.empty()) libraryDispatchContext.push_back(' ');
libraryDispatchContext += requirement.normalizedText;
}
}
const bool strictExecutionContract = args.value("strictExecutionContract", false);
const json executionHints = collectExecutionHints(normalized.requirements);
int escalateCount = 0;
int missingContractCount = 0;
int enrichedTaskCount = 0;
json tasksJson = json::array();
for (const auto& task : annotated) {
json taskJson = {
@@ -496,6 +538,31 @@
{"reasons", task.reasons}
};
json contract = buildExecutionContract(task, executionHints, strictExecutionContract);
if (libraryDispatchEnabled) {
std::string dispatchText = task.base.title;
if (!libraryDispatchContext.empty()) {
dispatchText += " ";
dispatchText += libraryDispatchContext;
}
auto enriched = libraryAnnotator.annotate(dispatchText, availableLibraries, capabilityLedger, symbolCatalog);
if (enriched.isEnriched()) {
contract["selectedLibrary"] = enriched.selectedLibrary;
contract["operationDomain"] = enriched.operationDomain;
contract["capabilityScore"] = enriched.capabilityScore;
contract["preferredAPIs"] = enriched.preferredAPIs;
contract["justification"] = enriched.justification;
json avoided = json::array();
for (const auto& item : enriched.avoidedLibraries) {
avoided.push_back({
{"name", item.name},
{"reason", item.reason},
{"score", item.score}
});
}
contract["avoidedLibraries"] = avoided;
++enrichedTaskCount;
}
}
taskJson["executionContract"] = contract;
taskJson["resourceLocks"] = inferResourceLocks(task);
if (strictExecutionContract && contract.value("executionSpecificityScore", 0) < 60) {
@@ -522,7 +589,12 @@
{"ambiguousRequirementCount", countAmbiguousRequirements(normalized.requirements)},
{"escalateCount", escalateCount},
{"strictExecutionContract", strictExecutionContract},
{"missingExecutionContractCount", missingContractCount}
{"missingExecutionContractCount", missingContractCount},
{"libraryDispatch", {
{"enabled", libraryDispatchEnabled},
{"availableLibraries", availableLibraries},
{"enrichedTaskCount", enrichedTaskCount}
}}
};
}

View File

@@ -279,6 +279,27 @@ if [[ "$SEMANTIC_PLANNING_BRIDGE" == "1" && "$SEMANTIC_INTAKE_AUGMENT" == "1" &&
else
MARKDOWN_CONTENT="$(cat "$INPUT_FILE")"
fi
AVAILABLE_LIBRARIES_JSON="${WSTONE_AVAILABLE_LIBRARIES_JSON:-}"
if [[ -z "$AVAILABLE_LIBRARIES_JSON" ]]; then
AVAILABLE_LIBRARIES_JSON="$(cat "$INPUT_FILE" | python3 -c '
import json, re, sys
text = sys.stdin.read()
libs = []
for line in text.splitlines():
m = re.search(r"available libraries\s*:\s*(.+)$", line, re.IGNORECASE)
if not m:
continue
for part in m.group(1).split(","):
name = part.strip().strip("`").strip()
if name and name not in libs:
libs.append(name)
print(json.dumps(libs))
')"
fi
if ! printf '%s' "$AVAILABLE_LIBRARIES_JSON" | jq -e 'type == "array"' >/dev/null 2>&1; then
echo "error: available libraries must be a JSON array (WSTONE_AVAILABLE_LIBRARIES_JSON)" >&2
exit 21
fi
INTAKE_ARGS="$(jq -nc --arg md "$MARKDOWN_CONTENT" '{markdown:$md}')"
INTAKE_RESP_RAW="$(call_tool "whetstone_architect_intake" "$INTAKE_ARGS")"
printf '%s\n' "$INTAKE_RESP_RAW" > "$OUT_DIR/01_intake_raw.ndjson.json"
@@ -402,8 +423,8 @@ else
NATIVE_RAW_TEMPLATE_CONTROL_PACK_JSON='{"enabled":false}'
fi
CONFLICTS="$(printf '%s' "$INTAKE_JSON" | jq '.conflicts // []')"
GEN_ARGS="$(jq -nc --argjson nr "$NORMALIZED_REQS" --argjson cf "$CONFLICTS" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,strictExecutionContract:($strict == "1")}')"
GEN_ARGS="$(jq -nc --argjson nr "$NORMALIZED_REQS" --argjson cf "$CONFLICTS" --argjson libs "$AVAILABLE_LIBRARIES_JSON" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,availableLibraries:$libs,strictExecutionContract:($strict == "1")}')"
GEN_RESP_RAW="$(call_tool "whetstone_generate_taskitems" "$GEN_ARGS")"
printf '%s\n' "$GEN_RESP_RAW" > "$OUT_DIR/02_generate_taskitems_raw.ndjson.json"
GEN_JSON="$(extract_tool_text_json "$GEN_RESP_RAW")"
@@ -426,8 +447,8 @@ if [[ "$NATIVE_DECOMP_RETRY" == "1" && "$native_task_count_initial" -lt "$NATIVE
ambiguous: false
}]
')"
RETRY_GEN_ARGS="$(jq -nc --argjson nr "$RETRY_REQS" --argjson cf "$CONFLICTS" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,strictExecutionContract:($strict == "1")}')"
RETRY_GEN_ARGS="$(jq -nc --argjson nr "$RETRY_REQS" --argjson cf "$CONFLICTS" --argjson libs "$AVAILABLE_LIBRARIES_JSON" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,availableLibraries:$libs,strictExecutionContract:($strict == "1")}')"
RETRY_GEN_RESP_RAW="$(call_tool "whetstone_generate_taskitems" "$RETRY_GEN_ARGS")"
printf '%s\n' "$RETRY_GEN_RESP_RAW" > "$OUT_DIR/02aa_generate_taskitems_retry_raw.ndjson.json"
RETRY_GEN_JSON="$(extract_tool_text_json "$RETRY_GEN_RESP_RAW")"
@@ -539,8 +560,8 @@ if [[ "$NATIVE_RAW_CANDIDATE_SEARCH" == "1" ]]; then
while [[ "$variant" -lt "$max_variants" && "$idx" -lt "$variant_total" ]]; do
req_bundle="$(printf '%s' "$VARIANTS_JSON" | jq ".variants[$idx]")"
cand_reqs="$(jq -nc --argjson base "$NORMALIZED_REQS" --argjson bundle "$req_bundle" '$base + $bundle')"
cand_args="$(jq -nc --argjson nr "$cand_reqs" --argjson cf "$CONFLICTS" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,strictExecutionContract:($strict == "1")}')"
cand_args="$(jq -nc --argjson nr "$cand_reqs" --argjson cf "$CONFLICTS" --argjson libs "$AVAILABLE_LIBRARIES_JSON" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,availableLibraries:$libs,strictExecutionContract:($strict == "1")}')"
cand_raw="$(call_tool "whetstone_generate_taskitems" "$cand_args")"
printf '%s\n' "$cand_raw" > "$OUT_DIR/02ae_candidate_${variant}_raw.ndjson.json"
cand_json="$(extract_tool_text_json "$cand_raw")"
@@ -603,8 +624,8 @@ if [[ "$NATIVE_RAW_CANDIDATE_SEARCH" == "1" ]]; then
--out-report "$OUT_DIR/02af_raw_adaptive_retry_requirements_report.json" >/dev/null
retry_top_gap_reqs="$(cat "$OUT_DIR/02af_raw_adaptive_retry_requirements.json")"
retry_cand_reqs="$(jq -nc --argjson base "$NORMALIZED_REQS" --argjson extra "$retry_top_gap_reqs" '$base + $extra')"
retry_cand_args="$(jq -nc --argjson nr "$retry_cand_reqs" --argjson cf "$CONFLICTS" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,strictExecutionContract:($strict == "1")}')"
retry_cand_args="$(jq -nc --argjson nr "$retry_cand_reqs" --argjson cf "$CONFLICTS" --argjson libs "$AVAILABLE_LIBRARIES_JSON" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,availableLibraries:$libs,strictExecutionContract:($strict == "1")}')"
retry_raw="$(call_tool "whetstone_generate_taskitems" "$retry_cand_args")"
printf '%s\n' "$retry_raw" > "$OUT_DIR/02af_raw_adaptive_retry_raw.ndjson.json"
retry_json="$(extract_tool_text_json "$retry_raw")"
@@ -791,8 +812,8 @@ if [[ "$NATIVE_MULTISHOT_DECOMP" == "1" ]]; then
'[($prof.profiles[]? | select(.id == $pid) | .required_execution_contract[]?) // empty]')"
shot_reqs="$(jq -nc --argjson base "$NORMALIZED_REQS" --argjson req "$req_i" '$base + [$req]')"
shot_args="$(jq -nc --argjson nr "$shot_reqs" --argjson cf "$CONFLICTS" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,strictExecutionContract:($strict == "1")}')"
shot_args="$(jq -nc --argjson nr "$shot_reqs" --argjson cf "$CONFLICTS" --argjson libs "$AVAILABLE_LIBRARIES_JSON" --arg strict "$STRICT_EXECUTION_CONTRACT" \
'{normalizedRequirements:$nr,conflicts:$cf,availableLibraries:$libs,strictExecutionContract:($strict == "1")}')"
shot_raw="$(call_tool "whetstone_generate_taskitems" "$shot_args")"
printf '%s\n' "$shot_raw" > "$OUT_DIR/02ac_generate_taskitems_multishot_${i}_raw.ndjson.json"
shot_json="$(extract_tool_text_json "$shot_raw")"