Skip to content

gridaco/grida / .agents/skills/grounding

grounding

Verified

Establish what is actually true and current for the surface you are about to change — not just search. Grounding = locate the authoritative source and reconcile sources that disagree (code vs doc, migration vs schema, memory vs current code, live vs archived), not take the first hit. Use before any grep/find/explore of the codebase or docs, when deciding which of several definitions is the real one, or when a doc or memory conflicts with the code. Covers the source-of-truth hierarchy, reconciliation discipline, scoped ripgrep, and the docs/tags.yml + docsearch.py index.

FileKindSize
SKILL.mdskill md4.4 KB
scripts/docsearch.pyexecutablescript10.8 KB

scripts/docsearch.py

View on GitHub
#!/usr/bin/env -S uv run --script
# /// script
# requires-python = ">=3.11"
# dependencies = ["pyyaml>=6"]
# ///
"""
docsearch — query Grida docs by YAML frontmatter ONLY.

Never reads the document body: it stops at the closing `---` fence, so
scanning all ~360 docs is one cheap pass. Use it to pick the 1-3 right
docs by tag/field before opening anything, instead of globbing or
grepping full files.

The docs root is auto-detected by walking up to `docs/tags.yml`, so this
runs from any cwd. Override with --root.

COMMANDS
  tags                       Tag vocabulary (docs/tags.yml) + live usage
                             counts + drift (defined-unused / used-undefined).
  find  [filters] [SUBDIR…]  Docs whose frontmatter matches ALL filters.
                             Optional SUBDIR args scope the scan (relative
                             to docs root, e.g. `wg/feat-fig reference`).
  show  PATH                 Parsed frontmatter for one file (path is
                             relative to cwd or docs root).

find FILTERS
  --tag T        require tag T   (repeatable; AND across tags by default)
  --any          combine --tag with OR instead of AND
  --field K=V    require frontmatter field K == V (repeatable, case-insens.;
                 list fields match if V is one of the items)
  --has K        require key K present in frontmatter (repeatable)
  --match RE     Python regex tested against the RAW frontmatter text only
  --json         machine-readable output

EXAMPLES
  docsearch tags
  docsearch tags --alpha
  docsearch find --tag figma --tag wg
  docsearch find --any --tag figma --tag forms wg
  docsearch find --has doc_tasks
  docsearch find --field draft=true
  docsearch find --match 'unlisted:\\s*true'
  docsearch show wg/feat-fig/glossary/fig.kiwi.md
"""

from __future__ import annotations

import argparse
import json
import re
import sys
from pathlib import Path

import yaml

MAX_FM_LINES = 300  # safety cap when a file has an opening `---` but no closer


def find_docs_root(override: str | None) -> Path:
    if override:
        p = Path(override).resolve()
        if (p / "tags.yml").exists():
            return p
        if (p / "docs" / "tags.yml").exists():
            return p / "docs"
        sys.exit(f"docsearch: no tags.yml under --root {p}")
    # walk up from cwd, then from this script's location
    for start in (Path.cwd(), Path(__file__).resolve().parent):
        cur = start
        for _ in range(40):
            if (cur / "docs" / "tags.yml").exists():
                return cur / "docs"
            if cur.name == "docs" and (cur / "tags.yml").exists():
                return cur
            if cur.parent == cur:
                break
            cur = cur.parent
    sys.exit("docsearch: could not locate docs/tags.yml (pass --root)")


def read_frontmatter(path: Path) -> tuple[dict | None, str]:
    """Return (parsed_dict_or_None, raw_block_text). Reads only the fence."""
    raw_lines: list[str] = []
    try:
        with path.open("r", encoding="utf-8-sig", errors="replace") as fh:
            first = fh.readline()
            if first.strip() != "---":
                return None, ""
            for _ in range(MAX_FM_LINES):
                line = fh.readline()
                if line == "":  # EOF before closer
                    return None, ""
                if line.strip() == "---":
                    raw = "".join(raw_lines)
                    try:
                        data = yaml.safe_load(raw)
                    except yaml.YAMLError:
                        return None, raw
                    return (data if isinstance(data, dict) else {}), raw
                raw_lines.append(line)
    except OSError:
        return None, ""
    return None, ""


def iter_docs(root: Path, subdirs: list[str]):
    root_resolved = root.resolve()
    if subdirs:
        bases = []
        for s in subdirs:
            if not (root_resolved / s).resolve().is_relative_to(root_resolved):
                sys.exit(f"docsearch: subdir escapes docs root: {s!r}")
            bases.append(root / s)
    else:
        bases = [root]
    seen: set[Path] = set()
    for base in bases:
        if not base.exists():
            sys.exit(f"docsearch: path not found: {base}")
        for path in sorted([*base.rglob("*.md"), *base.rglob("*.mdx")]):
            if path not in seen:
                seen.add(path)
                yield path


def as_list(v) -> list[str]:
    if v is None:
        return []
    if isinstance(v, (list, tuple)):
        return [str(x) for x in v]
    return [str(v)]


def field_matches(meta: dict, key: str, want: str) -> bool:
    if key not in meta:
        return False
    want_l = want.strip().lower()
    return any(x.strip().lower() == want_l for x in as_list(meta[key]))


def cmd_tags(root: Path, args) -> int:
    vocab_path = root / "tags.yml"
    vocab = yaml.safe_load(vocab_path.read_text(encoding="utf-8")) or {}

    counts: dict[str, int] = {}
    total = 0
    for path in iter_docs(root, []):
        meta, _ = read_frontmatter(path)
        if meta is None:
            continue
        total += 1
        for t in as_list(meta.get("tags")):
            counts[t] = counts.get(t, 0) + 1

    defined = set(vocab.keys())
    used = set(counts.keys())

    rows = []
    for key, spec in vocab.items():
        label = spec.get("label", key) if isinstance(spec, dict) else key
        desc = spec.get("description", "") if isinstance(spec, dict) else ""
        rows.append((key, label, desc, counts.get(key, 0)))

    if args.alpha:
        rows.sort(key=lambda r: r[0])
    else:
        rows.sort(key=lambda r: (-r[3], r[0]))

    if args.json:
        print(
            json.dumps(
                {
                    "docs_scanned": total,
                    "tags": [
                        {"key": k, "label": lb, "count": c, "description": d}
                        for (k, lb, d, c) in rows
                    ],
                    "defined_unused": sorted(defined - used),
                    "used_undefined": sorted(used - defined),
                },
                indent=2,
            )
        )
        return 0

    print(f"# {len(vocab)} tags defined in docs/tags.yml — {total} docs scanned\n")
    width = max((len(r[0]) for r in rows), default=4)
    for key, _label, desc, cnt in rows:
        mark = f"{cnt:>4}" if cnt else "   ·"
        print(f"{mark}  {key.ljust(width)}  {desc}")

    unused = sorted(defined - used)
    undef = sorted(used - defined)
    if unused:
        print(f"\n# defined but unused ({len(unused)}): {', '.join(unused)}")
    if undef:
        print(
            f"\n# WARNING used but NOT in tags.yml ({len(undef)}): "
            f"{', '.join(undef)}"
        )
    return 0


def cmd_find(root: Path, args) -> int:
    want_tags = [t.lower() for t in (args.tag or [])]
    fields = []
    for raw in args.field or []:
        if "=" not in raw:
            sys.exit(f"docsearch: --field expects K=V, got {raw!r}")
        k, v = raw.split("=", 1)
        fields.append((k.strip(), v))
    has_keys = args.has or []
    rx = None
    if args.match:
        try:
            rx = re.compile(args.match)
        except re.error as e:
            sys.exit(f"docsearch: invalid --match regex: {e}")

    results = []
    for path in iter_docs(root, args.subdir):
        meta, raw = read_frontmatter(path)
        if meta is None:
            continue
        if want_tags:
            doc_tags = {t.lower() for t in as_list(meta.get("tags"))}
            ok = (
                any(t in doc_tags for t in want_tags)
                if args.any
                else all(t in doc_tags for t in want_tags)
            )
            if not ok:
                continue
        if any(not field_matches(meta, k, v) for k, v in fields):
            continue
        if any(k not in meta for k in has_keys):
            continue
        if rx and not rx.search(raw):
            continue
        results.append((path, meta))

    rel = [
        (str(p.relative_to(root.parent) if root.parent in p.parents else p), m)
        for p, m in results
    ]

    if args.json:
        print(
            json.dumps(
                [
                    {
                        "path": pth,
                        "title": m.get("title"),
                        "tags": as_list(m.get("tags")),
                        "description": m.get("description"),
                    }
                    for pth, m in rel
                ],
                indent=2,
            )
        )
        return 0 if rel else 1

    for pth, m in rel:
        print(pth)
        if m.get("title"):
            print(f"  title: {m['title']}")
        tg = as_list(m.get("tags"))
        if tg:
            print(f"  tags:  {', '.join(tg)}")
        if m.get("description"):
            d = str(m["description"]).strip().replace("\n", " ")
            print(f"  desc:  {d[:140]}")
    print(f"\n# {len(rel)} match(es)")
    return 0 if rel else 1


def cmd_show(root: Path, args) -> int:
    p = Path(args.path)
    if not p.exists():
        p = root / args.path
    if not p.exists():
        p = root.parent / args.path
    if not p.exists():
        sys.exit(f"docsearch: file not found: {args.path}")
    meta, raw = read_frontmatter(p)
    if meta is None:
        print(f"{p}: no parseable frontmatter")
        return 1
    if args.json:
        print(json.dumps(meta, indent=2, default=str))
    else:
        print(f"# {p}\n")
        print(raw.rstrip())
    return 0


def main() -> int:
    ap = argparse.ArgumentParser(prog="docsearch", add_help=True)
    ap.add_argument("--root", help="docs dir (default: auto-detect via tags.yml)")
    sub = ap.add_subparsers(dest="cmd", required=True)

    t = sub.add_parser("tags", help="list tag vocabulary + usage")
    t.add_argument("--alpha", action="store_true", help="sort A-Z (default: by count)")
    t.add_argument("--json", action="store_true")

    f = sub.add_parser("find", help="find docs by frontmatter")
    f.add_argument("subdir", nargs="*", help="restrict to these subdirs of docs root")
    f.add_argument("--tag", action="append")
    f.add_argument("--any", action="store_true", help="OR the --tag filters")
    f.add_argument("--field", action="append", help="K=V frontmatter equality")
    f.add_argument("--has", action="append", help="frontmatter key must be present")
    f.add_argument("--match", help="regex over raw frontmatter text")
    f.add_argument("--json", action="store_true")

    s = sub.add_parser("show", help="print one file's frontmatter")
    s.add_argument("path")
    s.add_argument("--json", action="store_true")

    args = ap.parse_args()
    root = find_docs_root(args.root)

    if args.cmd == "tags":
        return cmd_tags(root, args)
    if args.cmd == "find":
        return cmd_find(root, args)
    if args.cmd == "show":
        return cmd_show(root, args)
    return 2


if __name__ == "__main__":
    raise SystemExit(main())
rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content.Stars: 197kUpdated Claude CodeCodexVerifiedInstall
claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file).Stars: 197kUpdated Claude CodeCodexUnverifiedInstall
doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.Stars: 197kUpdated Claude CodeCodexVerifiedInstall