Microsoft/bcquality
1
0
Fork 0
mirror of https://github.com/microsoft/BCQuality.git synced 2026-08-06 17:36:53 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions

R6 Tier-1: content-owned routing index generator + CI guard

Browse source 
Retire the orchestrator's hardcoded ~28-token signal catalog by compiling a
routing index from article front-matter. New tools/Build-RoutingIndex.ps1 scans
all knowledge layers and emits routing-index.json (signal-token -> domain +
backing articles), seeded from the migrated legacy catalog (routing-seed.json)
so recall is never below today. Optional per-article 'signals:' front-matter
(validator rule R24) is the authored precision path; domain-normalization
reconciles front-matter/orchestrator/feedback domain vocabularies.

Artifact is a runtime build (gitignored), orchestrator-facing only (not the
lean agent-replayed knowledge index). CI guard Test-RoutingIndex.ps1 asserts
determinism, seed recall floor, no orphaned signals, and normalization
completeness; wired into the index workflow.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This commit is contained in:
dayland 2026-07-13 11:20:30 +01:00
parent 4119417ce4
commit 6ed56b95b7
8 changed files with 722 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

32
.github/scripts/validate_frontmatter.py vendored
View file

@ -34,6 +34,11 @@ KNOWLEDGE_REQUIRED_KEYS = {
"bc-version", "domain", "keywords", "technologies",
"countries", "application-area",
}
# Optional knowledge keys (do not trigger the R02 closed-key-set error).
# `signals`: OPTIONAL routing-signal declarations consumed by the routing index
# (tools/Build-RoutingIndex.ps1). Back-compatible — articles that omit it are
# routed from the seed catalog + keyword derivation. See tools/routing-index.md.
KNOWLEDGE_OPTIONAL_KEYS = {"signals"}
ACTION_SKILL_REQUIRED_KEYS = {
"kind", "id", "version", "title", "description", "inputs", "outputs",
}
@ -202,7 +207,7 @@ def validate_knowledge(path: Path, parsed: Parsed, report: Report) -> None:
# R02 required keys, no extras, none empty
missing = KNOWLEDGE_REQUIRED_KEYS - fm.keys()
extras = fm.keys() - KNOWLEDGE_REQUIRED_KEYS
extras = fm.keys() - KNOWLEDGE_REQUIRED_KEYS - KNOWLEDGE_OPTIONAL_KEYS
if missing:
report.error(path, "R02", f"missing required frontmatter keys: {sorted(missing)}", 1)
if extras:
@ -212,6 +217,31 @@ def validate_knowledge(path: Path, parsed: Parsed, report: Report) -> None:
if v is None or v == "" or v == []:
report.error(path, "R02", f"frontmatter key '{k}' must not be empty", 1)
# R24 optional routing `signals` block. Each entry is either a bare token
# string or a mapping with a required 'token' and optional 'pattern'/'domain'
# (all non-empty strings). Keeps the routing-index generator's input honest.
if "signals" in fm:
sigs = fm["signals"]
if not isinstance(sigs, list) or not sigs:
report.error(path, "R24", "signals must be a non-empty list", 1)
else:
for entry in sigs:
if isinstance(entry, str):
if not entry.strip():
report.error(path, "R24", "signals token string must not be empty", 1)
elif isinstance(entry, dict):
tok = entry.get("token")
if not isinstance(tok, str) or not tok.strip():
report.error(path, "R24", "signals entry must have a non-empty 'token'", 1)
for opt in ("pattern", "domain"):
if opt in entry and (not isinstance(entry[opt], str) or not entry[opt].strip()):
report.error(path, "R24", f"signals '{opt}' must be a non-empty string", 1)
unknown = set(entry.keys()) - {"token", "pattern", "domain"}
if unknown:
report.error(path, "R24", f"signals entry has unknown keys: {sorted(unknown)}", 1)
else:
report.error(path, "R24", "signals entry must be a string or a mapping", 1)
# R03 bc-version
if "bc-version" in fm:
_, err = expand_bc_version(fm["bc-version"])