Skip to content

zrwangems/dsh-zrwang-plugin / ars/skills/academic-pipeline

academic-pipeline

Flagged

学术研究总编排:研究→写作→诚信检查→评审→返修→复审→终检→定稿的十阶段流水线,每阶段停在用户确认点,含材料护照与 7 类 AI 研究失败模式检查。Triggers: academic pipeline, research to paper, full paper workflow, end-to-end paper | 完整论文流程, 从研究到论文, 端到端写论文, 论文流水线。

FileKindSize
SKILL.mdskill md58.4 KB
_shared/ARCHITECTURE.mdmarkdown51.7 KB
_shared/MODE_REGISTRY.mdmarkdown6.7 KB
_shared/agents/compliance_agent.mdmarkdown9.4 KB
_shared/artifact_reproducibility_pattern.mdmarkdown9.8 KB
_shared/benchmark_report.schema.jsonconfig2.7 KB
_shared/benchmark_report_pattern.mdmarkdown8.8 KB
_shared/bibliographic_integrity_signals.mdmarkdown8.1 KB
_shared/collaboration_depth_rubric.mdmarkdown11.5 KB
_shared/compliance_checkpoint_protocol.mdmarkdown8.9 KB
_shared/compliance_report.schema.jsonconfig8.1 KB
_shared/contracts/README.mdmarkdown54.7 KB
_shared/contracts/activity/adjudication_activity_input.schema.jsonconfig18.5 KB
_shared/contracts/activity/adjudication_activity_store.schema.jsonconfig18.7 KB
_shared/contracts/audit/audit_jsonl.schema.jsonconfig5.4 KB
_shared/contracts/audit/audit_sidecar.schema.jsonconfig8.4 KB
_shared/contracts/audit/audit_verdict.schema.jsonconfig5.4 KB
_shared/contracts/audit/cross_document_consistency_advisory.schema.jsonconfig20.4 KB
_shared/contracts/audit/cross_document_consistency_advisory_draft.schema.jsonconfig23.4 KB
_shared/contracts/audit/cross_document_source_manifest.schema.jsonconfig6.3 KB
_shared/contracts/audit/tortured_phrase_advisory.schema.jsonconfig30.9 KB
_shared/contracts/audit/tortured_phrase_snapshot.schema.jsonconfig4.9 KB
_shared/contracts/audit/tortured_phrase_snapshot_manifest.schema.jsonconfig7.3 KB
_shared/contracts/capability/stage_capability_matrix.jsonconfig23.5 KB
_shared/contracts/claim_standing/candidate_ledger.schema.jsonconfig23.1 KB
_shared/contracts/claim_standing/query_plan.schema.jsonconfig12.7 KB
_shared/contracts/claim_standing/query_plan_v1_1.schema.jsonconfig20.3 KB
_shared/contracts/claim_standing/retrieval_input.schema.jsonconfig14.4 KB
_shared/contracts/claim_standing/stance_record.schema.jsonconfig11.7 KB
_shared/contracts/claim_standing/transmission_ledger.schema.jsonconfig7.6 KB
_shared/contracts/cross_model/codex_citation_receipt.schema.jsonconfig4.7 KB
_shared/contracts/cross_model/codex_citation_request.schema.jsonconfig986 B
_shared/contracts/cross_model/promotion_bakeoff_sealed_commitment.schema.jsonconfig1.5 KB
_shared/contracts/cross_model/promotion_bakeoff_sealed_reveal.schema.jsonconfig1.5 KB
_shared/contracts/degradation_registry.jsonconfig42.5 KB
_shared/contracts/evaluator/full.jsonconfig6.3 KB
_shared/contracts/evidence/claim_registry.schema.jsonconfig2.3 KB
_shared/contracts/evidence/claim_registry_coverage_report.schema.jsonconfig4.8 KB
_shared/contracts/evidence/evidence_row.schema.jsonconfig14.3 KB
_shared/contracts/evidence/evidence_row_v1_1.schema.jsonconfig11.4 KB
_shared/contracts/evidence/evidence_row_v1_2.schema.jsonconfig21.8 KB
_shared/contracts/evidence/evidence_row_v1_3.schema.jsonconfig10.7 KB
_shared/contracts/human_subjects/authority_profile_registry.schema.jsonconfig16.7 KB
_shared/contracts/human_subjects/committee_correspondence.schema.jsonconfig9.9 KB
_shared/contracts/human_subjects/content_coverage_advisory.schema.jsonconfig19.9 KB
_shared/contracts/human_subjects/irb_context_record.schema.jsonconfig12.1 KB
_shared/contracts/human_subjects/resolved_authority_context.schema.jsonconfig11.8 KB
_shared/contracts/human_subjects/review_pathway_rule_trace.schema.jsonconfig9.5 KB
_shared/contracts/human_subjects/review_pathway_trace_request.schema.jsonconfig4.8 KB
_shared/contracts/human_subjects/submission_packet_inventory.schema.jsonconfig8.5 KB
_shared/contracts/human_subjects/submission_packet_manifest.schema.jsonconfig32.2 KB
_shared/contracts/passport/audit_artifact_entry.schema.jsonconfig10.8 KB
_shared/contracts/passport/bibliographic_integrity_signal.schema.jsonconfig48.9 KB
_shared/contracts/passport/citation_provenance.schema.jsonconfig4.8 KB
_shared/contracts/passport/citation_verification_summary.schema.jsonconfig8.2 KB
_shared/contracts/passport/claim_audit_result.schema.jsonconfig7.8 KB
_shared/contracts/passport/claim_drift.schema.jsonconfig3.8 KB
_shared/contracts/passport/claim_intent_manifest.schema.jsonconfig6.6 KB
_shared/contracts/passport/constraint_violation.schema.jsonconfig3.6 KB
_shared/contracts/passport/excluded_source_entry.schema.jsonconfig1.7 KB
_shared/contracts/passport/experiment_alignment_result.schema.jsonconfig4.8 KB
_shared/contracts/passport/experiment_provenance_entry.schema.jsonconfig10.0 KB
_shared/contracts/passport/human_read_log.schema.jsonconfig3.8 KB
_shared/contracts/passport/inquiry_ledger_ref.schema.jsonconfig881 B
_shared/contracts/passport/literature_corpus_entry.schema.jsonconfig30.3 KB
_shared/contracts/passport/preregistration_artifact.schema.jsonconfig4.3 KB
_shared/contracts/passport/rejection_log.schema.jsonconfig3.3 KB
_shared/contracts/passport/reset_ledger_entry.schema.jsonconfig7.1 KB
_shared/contracts/passport/run_ledger.schema.jsonconfig8.0 KB
_shared/contracts/passport/standing_constraint_entry.schema.jsonconfig2.6 KB
_shared/contracts/passport/temporal_audit_results.schema.jsonconfig8.8 KB
_shared/contracts/passport/terminal_policies.schema.jsonconfig8.1 KB
_shared/contracts/passport/timeline.schema.jsonconfig4.7 KB
_shared/contracts/passport/uncited_assertion.schema.jsonconfig3.3 KB
_shared/contracts/passport/uncited_audit_failure.schema.jsonconfig4.4 KB
_shared/contracts/passport/user_attested_read_resolution.schema.jsonconfig3.2 KB
_shared/contracts/passport/version_records.schema.jsonconfig4.9 KB
_shared/contracts/patch/block_manifest.schema.jsonconfig2.1 KB
_shared/contracts/patch/legacy/v1_0/revision_patch.schema.jsonconfig5.3 KB
_shared/contracts/patch/revision_patch.schema.jsonconfig8.9 KB
_shared/contracts/pdf/pdf_content_classifier_diagnostic.schema.jsonconfig1.2 KB
_shared/contracts/pdf/pdf_content_classifier_worker.schema.jsonconfig2.2 KB
_shared/contracts/pdf/pdf_read_preflight.schema.jsonconfig6.8 KB
_shared/contracts/re_review/input_manifest.schema.jsonconfig7.6 KB
_shared/contracts/re_review/legacy/v1_0/input_manifest.schema.jsonconfig6.2 KB
_shared/contracts/re_review/legacy/v1_0/precommitment.schema.jsonconfig8.9 KB
_shared/contracts/re_review/legacy/v1_0/traceability.schema.jsonconfig34.2 KB
_shared/contracts/re_review/legacy/v1_0/verdict_record.schema.jsonconfig10.5 KB
_shared/contracts/re_review/precommitment.schema.jsonconfig8.6 KB
_shared/contracts/re_review/traceability.schema.jsonconfig37.7 KB
_shared/contracts/re_review/verdict_record.schema.jsonconfig10.5 KB
_shared/contracts/research_workflow/inquiry_branch_ledger.schema.jsonconfig17.0 KB
_shared/contracts/research_workflow/research_workflow_profile.schema.jsonconfig7.2 KB
_shared/contracts/research_workflow/research_workflow_profile_selection_receipt.schema.jsonconfig3.9 KB
_shared/contracts/review_target/constructive_review_findings.schema.jsonconfig7.0 KB
_shared/contracts/review_target/criteria_registry.schema.jsonconfig4.0 KB
_shared/contracts/review_target/review_criteria_binding_manifest.schema.jsonconfig7.9 KB
_shared/contracts/review_target/review_criteria_source_receipt.schema.jsonconfig5.6 KB
_shared/contracts/review_target/review_target_context.schema.jsonconfig4.5 KB
_shared/contracts/review_target/review_target_declaration.schema.jsonconfig4.6 KB
_shared/contracts/reviewer/full.jsonconfig3.7 KB
_shared/contracts/reviewer/methodology_focus.jsonconfig2.2 KB
_shared/contracts/reviewer/review_panel_provenance.schema.jsonconfig7.8 KB
_shared/contracts/reviewer/review_panel_provenance_carrier.schema.jsonconfig4.1 KB
_shared/contracts/reviewer/review_panel_provenance_input.schema.jsonconfig3.5 KB
_shared/contracts/revision/author_adjudication.schema.jsonconfig7.8 KB
_shared/contracts/revision/author_adjudication_input.schema.jsonconfig1.2 KB
_shared/contracts/revision/claim_strength_drift_disposition.schema.jsonconfig3.6 KB
_shared/contracts/revision/claim_strength_drift_disposition_input.schema.jsonconfig3.1 KB
_shared/contracts/revision/claim_strength_drift_findings.schema.jsonconfig4.2 KB
_shared/contracts/revision/claim_surface_manifest.schema.jsonconfig3.3 KB
_shared/contracts/revision/integrity_correction_authorization.schema.jsonconfig4.3 KB
_shared/contracts/revision/integrity_correction_authorization_input.schema.jsonconfig1.5 KB
_shared/contracts/revision/integrity_correction_list.schema.jsonconfig1.8 KB
_shared/contracts/revision/integrity_pass_receipt.schema.jsonconfig809 B
_shared/contracts/revision/revision_evidence_bundle.schema.jsonconfig4.9 KB
_shared/contracts/revision/revision_roadmap.schema.jsonconfig11.2 KB
_shared/contracts/submission/format_profile.example.yamlconfig971 B
_shared/contracts/submission/format_profile.schema.jsonconfig4.6 KB
_shared/contracts/submission/submission_verification_report.schema.jsonconfig10.3 KB
_shared/contracts/submission/venue_profile.schema.jsonconfig4.1 KB
_shared/contracts/writer/full.jsonconfig3.6 KB
_shared/cross_model_verification.mdmarkdown83.3 KB
_shared/description-original.mdmarkdown904 B
_shared/dsh-runtime.mdmarkdown1.5 KB
_shared/evals_lift_report.schema.jsonconfig5.0 KB
_shared/ground_truth_isolation_pattern.mdmarkdown16.0 KB
_shared/handoff_schemas.mdmarkdown96.9 KB
_shared/human_subjects_authority_registry.jsonconfig49.5 KB
_shared/mode_spectrum.mdmarkdown5.2 KB
_shared/model_tiering.mdmarkdown12.6 KB
_shared/output_language_pair.mdmarkdown7.7 KB
_shared/policy_data/nature_policy.mdmarkdown3.9 KB
_shared/prisma_trAIce_protocol.mdmarkdown10.2 KB
_shared/raise_framework.mdmarkdown7.5 KB
_shared/references/authority_content_coverage_advisory_protocol.mdmarkdown14.0 KB
_shared/references/claim_standing_candidate_ledger_protocol.mdmarkdown19.2 KB
_shared/references/claim_strength_ladder.mdmarkdown5.5 KB
_shared/references/cross_document_consistency_advisory_protocol.mdmarkdown13.5 KB
_shared/references/evidence_row_protocol.mdmarkdown14.4 KB
_shared/references/firm_rules.mdmarkdown11.0 KB
_shared/references/human_subjects_authority_protocol.mdmarkdown13.1 KB
_shared/references/intent_clarification_protocol.mdmarkdown11.9 KB
_shared/references/irb_terminology_glossary.mdmarkdown12.4 KB
_shared/references/per_source_method_weaknesses.mdmarkdown4.2 KB
_shared/references/protected_hedging_phrases.mdmarkdown8.9 KB
_shared/references/psychometric_terminology_glossary.mdmarkdown8.7 KB
_shared/references/review_criteria_consumer_protocol.mdmarkdown10.3 KB
_shared/references/review_form_note.mdmarkdown11.3 KB
_shared/references/review_pathway_rule_trace_protocol.mdmarkdown8.3 KB
_shared/references/routing_core.mdmarkdown5.7 KB
_shared/references/submission_packet_manifest_protocol.mdmarkdown13.6 KB
_shared/references/word_count_conventions.mdmarkdown9.0 KB
_shared/research_workflow_profiles/field_general.jsonconfig1.2 KB
_shared/review_criteria_registry.jsonconfig11.5 KB
_shared/review_criteria_sources/msr-2027-technical-papers.2026-08-24.jsonconfig2.0 KB
_shared/review_criteria_sources/sigsoft-empirical-standards.2026-08-24.jsonconfig1.4 KB
_shared/sprint_contract.schema.jsonconfig19.0 KB
_shared/style_calibration_protocol.mdmarkdown6.9 KB
_shared/templates/codex_audit_multifile_template.mdmarkdown16.7 KB
agents/claim_ref_alignment_audit_agent.mdmarkdown45.0 KB
agents/collaboration_depth_agent.mdmarkdown10.4 KB
agents/integrity_verification_agent.mdmarkdown75.2 KB
agents/pipeline_orchestrator_agent.mdmarkdown175.5 KB
agents/state_tracker_agent.mdmarkdown27.4 KB
examples/full_pipeline_example.mdmarkdown17.7 KB
examples/integrity_failure_recovery.mdmarkdown26.8 KB
examples/mid_entry_example.mdmarkdown13.4 KB
references/adapters/.gitkeepother0 B
references/adapters/overview.mdmarkdown10.0 KB
references/ai_research_failure_modes.mdmarkdown16.3 KB
references/changelog.mdmarkdown6.4 KB
references/claim_audit_calibration_protocol.mdmarkdown14.9 KB
references/claim_verification_protocol.mdmarkdown31.1 KB
references/external_review_protocol.mdmarkdown6.8 KB
references/integrity_review_protocol.mdmarkdown7.8 KB
references/literature_corpus_consumers.mdmarkdown12.3 KB
references/mode_advisor.mdmarkdown8.7 KB
references/passport_as_reset_boundary.mdmarkdown19.3 KB
references/pipeline_state_machine.mdmarkdown29.0 KB
references/plagiarism_detection_protocol.mdmarkdown14.1 KB
references/process_summary_protocol.mdmarkdown16.4 KB
references/progress_dashboard_template.mdmarkdown2.1 KB
references/reinforcement_content.mdmarkdown2.0 KB
references/reproducibility_audit.mdmarkdown3.2 KB
references/score_trajectory_protocol.mdmarkdown4.5 KB
references/team_collaboration_protocol.mdmarkdown10.7 KB
references/two_stage_review_protocol.mdmarkdown3.6 KB
scripts/_claim_audit_constants.pyscript13.7 KB
scripts/_markdown_lint_util.pyscript9.0 KB
scripts/_next_verified_at_ms.pyscript6.5 KB
scripts/_skill_lint.pyscript13.6 KB
scripts/ars_anchorize_draft.pyscript5.8 KB
scripts/ars_apply_revision_patch.pyscript36.2 KB
scripts/ars_cache_invalidate.pyscript1.3 KB
scripts/ars_mark_read.pyscript18.2 KB
scripts/ars_write_scope_guard.pyscript30.9 KB
scripts/audit_snapshot.pyscript23.5 KB
scripts/bibliographic_integrity_signals.pyscript32.2 KB
scripts/build_claim_standing_candidate_ledger.pyscript54.3 KB

scripts/_claim_audit_constants.py

View on GitHub
"""Shared constants for v3.8 claim-faithfulness audit.

Single source of truth for the literals + regexes that appear in BOTH the
lint (`check_claim_audit_consistency.py`) and the pipeline runtime
(`claim_audit_pipeline.py`). Re-declaring these in both places opens a
drift hole — a spec bump that updates the lint without updating the
runtime would change one side silently. Tests cover both call sites; the
shared import binds them.

See docs/design/2026-05-15-issue-103-claim-alignment-audit-spec.md §3.1
(matrix + INV catalogue) and §4 step 3 (sampling) for canonical
definitions.
"""
from __future__ import annotations

import re

# Canonical sentinel for the MANIFEST-MISSING fallback path (spec §3.1 INV-15).
SENTINEL_MANIFEST_ID = "M-0000-00-00T00:00:00Z-0000"

# INV-6 canonical rationale prefix (v3.7.3 R-L3-1-A firm rule).
INV6_RATIONALE_PREFIX = "v3.7.3 R-L3-1-A violation"

# INV-14 audit-tool-failure rationale fault-class tags.
INV14_FAULT_CLASS_TAGS: tuple[str, ...] = (
    "judge_timeout",
    "judge_api_error",
    "judge_parse_error",
    "cache_corruption",
    "retrieval_api_error",
    "retrieval_timeout",
    "retrieval_network_error",
)

# Sampling strategy literal (S-INV schema constant).
SAMPLING_STRATEGY = "stratified_buckets_v1"

# Sub-claim breakdown sub_verdict enum (claim_audit_result.schema.json #213).
# The "non-SUPPORTED" set is the valid OPPOSING verdicts only — a missing or
# out-of-enum sub_verdict must NOT count as non-SUPPORTED, or a degenerate
# breakdown `[SUPPORTED, <missing>]` would masquerade as true-partial.
SUBCLAIM_VERDICTS: frozenset[str] = frozenset({"SUPPORTED", "UNSUPPORTED", "AMBIGUOUS"})
SUBCLAIM_NON_SUPPORTED: frozenset[str] = frozenset({"UNSUPPORTED", "AMBIGUOUS"})


def is_true_partial_breakdown(breakdown: object) -> bool:
    """True iff `breakdown` is a well-formed true-partial decomposition (#213).

    Single source of truth for the INV-19 true-partial test, shared by the lint
    (`check_claim_audit_consistency.py`), the runtime
    (`claim_audit_pipeline.py` PARTIAL normalization + judge-output validation),
    and the calibration subset metric (`claim_audit_calibration.py`). A list of
    >=2 dict items whose sub_verdicts include >=1 SUPPORTED AND >=1 valid
    non-SUPPORTED ({UNSUPPORTED, AMBIGUOUS}). A missing / out-of-enum sub_verdict
    is NOT counted as non-SUPPORTED.

    NOTE: this is the *content-shape* gate only. The lint additionally pins the
    enclosing row's judgment/defect_stage (INV-19) and the calibration subset
    metric additionally matches the breakdown against each fixture's expected
    sub-claims — neither of those belongs here.
    """
    if not isinstance(breakdown, list) or len(breakdown) < 2:
        return False
    verdicts = [item.get("sub_verdict") for item in breakdown if isinstance(item, dict)]
    has_supported = any(v == "SUPPORTED" for v in verdicts)
    has_non_supported = any(v in SUBCLAIM_NON_SUPPORTED for v in verdicts)
    return has_supported and has_non_supported


def _is_schema_shaped_item(item: object) -> bool:
    """True iff a breakdown item satisfies the schema item shape (#213).

    Each item MUST be a dict with a non-empty-string `sub_claim_text` and a
    `sub_verdict` in the closed enum. The runtime needs this BEFORE it copies an
    item onto an emitted row — `is_true_partial_breakdown` only checks the verdict
    *mix*, so a degenerate item like `{"sub_verdict": "UNSUPPORTED"}` (no text)
    passes the mix gate but would emit `sub_claim_text: None`, a schema-invalid
    completed row (ship-gate round-2 finding).

    Validates the fields the runtime COPIES onto the row against the
    claim_audit_result.schema.json item shape: non-empty string sub_claim_text
    (<=1000), sub_verdict in the enum, and evidence_pointer — if present — a
    string (<=1000) or null. Extra keys are not rejected here because the runtime
    copy keeps only these three; but the copied fields' TYPES must be valid or a
    wrong-typed evidence_pointer (e.g. a number) would reach a completed row and
    violate the schema (ship-gate round-3 finding).
    """
    if not isinstance(item, dict):
        return False
    text = item.get("sub_claim_text")
    if not isinstance(text, str) or not text.strip() or len(text) > 1000:
        return False
    if item.get("sub_verdict") not in SUBCLAIM_VERDICTS:
        return False
    if "evidence_pointer" in item:
        ep = item["evidence_pointer"]
        if ep is not None and (not isinstance(ep, str) or len(ep) > 1000):
            return False
    return True


def is_emittable_partial_breakdown(breakdown: object) -> bool:
    """True iff `breakdown` is true-partial AND every item is schema-shaped (#213).

    The runtime validation gate before a PARTIAL is normalized onto a *completed*
    row: it must be a genuine partial (`is_true_partial_breakdown`) AND every item
    must carry a non-empty sub_claim_text + valid sub_verdict, so the copied row
    satisfies the item schema. A breakdown that is true-partial by mix but has a
    malformed item is a judge parse failure, NOT a completed row.
    """
    if not is_true_partial_breakdown(breakdown):
        return False
    return all(_is_schema_shaped_item(item) for item in breakdown)

# rule_version literals for v3.8.0 release. Future revisions bump the literal
# and require re-lint per spec §3.3 / §3.4 / §3.5.
UNCITED_RULE_VERSION = "D4-c-v1"
DRIFT_RULE_VERSION = "D4-a-v1"
# v3.8.2 / #118 — uncited_audit_failure rule_version literal (§3.6).
# Distinct prefix from D4-c-v1 (uncited_assertion D4-c detector) and
# D4-a-v1 (constraint_violation) so the lint can route by literal.
UAF_RULE_VERSION = "D4-c-v1-uaf-v1"

# #361 — judge-prompt version, a HUMAN-READABLE LABEL ONLY. The judge prompt
# text lives in academic-pipeline/agents/claim_ref_alignment_audit_agent.md
# (### Step 5 — Judge invocation) and is supplied to the pipeline via an injected
# judge_fn, so the pipeline cannot hash the prompt text itself. This literal is a
# friendly name for the current prompt revision (e.g. the #213 Step-0
# sub-claim-decomposition prompt); BUMP IT whenever the judge prompt changes so
# logs/diffs read clearly. It is NO LONGER the judge-cache-key source of truth —
# the cache key tracks JUDGE_PROMPT_SHA256 below (the prompt fingerprint), which
# the lint keeps in lockstep with the prompt text. Decoupling the two means a
# forgotten label bump can never leave stale cache entries valid.
JUDGE_PROMPT_VERSION = "step0-decomp-v2-data-boundary"

# #361 backstop: SHA-256 of the canonical judge-prompt section (the text between
# the JUDGE-PROMPT-CANONICAL-START/END markers in
# academic-pipeline/agents/claim_ref_alignment_audit_agent.md, stripped).
# scripts/check_judge_prompt_version.py recomputes this hash and fails CI if it
# drifts — forcing a human to update this hash whenever the prompt text changes.
# This hash DOUBLES AS the judge-cache-key prompt component (the single source of
# truth for cache invalidation): claim_audit_pipeline.py defaults
# prompt_version to this value, so because the lint pins the hash to the prompt
# text, any prompt edit automatically changes the cache key and invalidates stale
# entries — no reliance on a separate human-readable label bump.
JUDGE_PROMPT_SHA256 = "bb67174bb664bc7a668de790012562cdc4f1410fbc35847531732699792d0ed2"

# Constraint id parse rules (spec §3.2 + INV-17 canonical form).
RE_NC_CONSTRAINT = re.compile(r"^NC-C([0-9]{3,})-([0-9]+)$")
RE_MNC_CONSTRAINT = re.compile(r"^MNC-([0-9]+)$")
RE_CLAIM_ID = re.compile(r"^C-([0-9]{3,})$")

# Schema rejects this pattern, but for malformed-on-purpose fixtures the lint
# surfaces INV-17 explicitly before schema validation runs.
RE_NC_INNER_HYPHEN = re.compile(r"^NC-C-[0-9]+-[0-9]+$")

# ---------------------------------------------------------------------------
# D4-c uncited-assertion detector constants (spec §"Uncited-assertion
# detector (D4-c)" in claim_ref_alignment_audit_agent.md).
#
# Centralised here so pipeline runtime, lint, and detector share one source
# of truth — a spec bump touches one literal, not three.
# ---------------------------------------------------------------------------

# Condition 1: empirical-claim verbs (case-insensitive whole-word match).
# Spec list: showed, demonstrated, observed, proved, confirmed.
UNCITED_EMPIRICAL_VERBS: frozenset[str] = frozenset(
    {"showed", "demonstrated", "observed", "proved", "confirmed"}
)

# Condition 1: fuzzy English quantifier words (case-insensitive whole-word).
# Spec list: most, several, two-thirds. Kept literal; numerical / percent
# quantifiers are caught by RE_NUMERIC_QUANTIFIER below.
UNCITED_FUZZY_QUANTIFIERS: frozenset[str] = frozenset(
    {"most", "several", "two-thirds"}
)

# Condition 1: numerical quantifier regex. Spec line 250 lists three numeric
# classes — `numbers / percentages / explicit quantifiers (50%, 67 of 100)`.
# All three fire D4-c condition 1; the detector then applies a guard pass
# (RE_NUMERIC_QUANTIFIER_GUARD below) that rejects matches whose surrounding
# context proves them to be years, version triples, or section numbers
# instead of quantifiers. Splitting into match-broadly + guard-narrowly is
# easier to read and to test than stuffing every exclusion into a single
# regex's negative lookaheads. The detector concatenates the matched
# substring into trigger_tokens verbatim so the schema's minItems=1
# invariant holds.
RE_NUMERIC_QUANTIFIER = re.compile(
    # Order matters: longest-prefix-first so percent and "N of M" bind
    # before the bare-number branch swallows the leading digits.
    r"\b\d+(?:\.\d+)?%"                  # percent quantifier
    r"|\b\d+(?:\.\d+)?\s+of\s+\d+\b"     # "N of M" quantifier idiom
    r"|\b\d+(?:\.\d+)*\b"                # bare number, possibly dotted
                                          # (3+ segments routed to guard
                                          # as version/section)
)

# Condition 1 guard: rejects bare-number matches whose surrounding context
# identifies them as years, version triples, or section numbers — none of
# those are quantitative claims, and treating them as such produced
# false-positive LOW-WARN advisories before the guard landed (see codex
# R1 P1-3). Applied AFTER RE_NUMERIC_QUANTIFIER to the matched substring +
# its character offsets in the sentence; bare-number matches that satisfy
# any guard branch are dropped, percent and `N of M` matches always pass
# through.
#
# Guard branches:
#   1. Standalone 4-digit year in plausible academic range (1900-2099).
#   2. Version triple `X.Y.Z` (dotted form with 3+ segments — the broad
#      regex captures only the first two segments, so we re-scan).
#   3. Dotted section number `X.Y[.Z…]` (treated as section ref, not
#      quantifier). Distinguished from version by `section` / `§` /
#      `chapter` / `figure` / `table` cue word within a 24-char left
#      window, OR by `v` immediately preceding (version literal).
RE_BARE_NUMERIC_YEAR = re.compile(r"^(19|20)\d{2}$")
RE_DOTTED_TRIPLE_OR_MORE = re.compile(r"^\d+(?:\.\d+){2,}$")
RE_DOTTED_PAIR = re.compile(r"^\d+\.\d+$")
RE_SECTION_CUE = re.compile(
    r"(?:section|chapter|figure|table|fig\.|tbl\.|step|appendix|§)\s*$",
    re.IGNORECASE,
)
RE_VERSION_PREFIX = re.compile(r"v\s*$", re.IGNORECASE)
# Catches the case where Python's `\b` fails between a letter and a digit
# (both are \w characters) — e.g. `v3.7.3` has no \b between `v` and `3`,
# so RE_NUMERIC_QUANTIFIER starts matching from the SECOND segment (`7.3`)
# and the guard never sees the version-triple shape. This pattern detects
# a digit-then-dot prefix immediately attached to the left of the match.
# Requires exactly one digit-run + `.` ending the window — combined with
# the surrounding match (also dotted or bare number), this signals a
# multi-segment dotted form that should be treated as a version/section
# reference rather than a quantifier.
RE_NUMERIC_LEFT_ATTACHED = re.compile(r"\d+\.$")

# Condition 2: three-layer-citation ref-marker probe. Presence probe
# WITHIN the v3.7.3 ref-marker namespace — accepts any `<!--ref:...-->`
# shape where the slug payload begins with a non-whitespace character,
# rejects HTML comments that happen to start with `ref:` but use it as
# a label rather than a citation marker (e.g.
# `<!-- ref: $internal.notebook.cell -->`).
#
# Iteration history:
#   R0 `[^-]+`                                — rejected hyphenated slugs.
#   R1 `[A-Za-z][A-Za-z0-9_:-]* + 0-2 tokens` — rejected digit-leading
#                                                 slugs, plus-sign slugs,
#                                                 and 3+ status tokens.
#   R2 `[^>]*?`                                — too broad; matched
#                                                 `<!-- ref: $analysis -->`
#                                                 and similar code/internal
#                                                 ref comments (codex
#                                                 R3 P2-NEW-A).
#   R3 `[^\s>][^>]*?`                          — current. Slug must begin
#                                                 with a non-whitespace
#                                                 non-`>` character; the
#                                                 v3.7.3 strict validator
#                                                 in
#                                                 scripts/check_v3_7_3_three_layer_citation.py
#                                                 catches any remaining
#                                                 shape errors.
RE_REF_MARKER = re.compile(r"<!--\s*ref:[^\s>][^>]*?-->")

# Condition 3: definitional-phrase substrings (case-insensitive). Spec list:
# `refers to`, `is defined as`, `we define`, `for the purposes of`.
UNCITED_DEFINITION_PHRASES: tuple[str, ...] = (
    "refers to",
    "is defined as",
    "we define",
    "for the purposes of",
)
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 322kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall