Quickstart: attune-rag editor

Lint a template file and see its diagnostics in under a minute:

from attune_rag.editor import lint_template

text = open("templates/my_template.yaml").read()
diagnostics = lint_template(text, rel_path="templates/my_template.yaml", corpus=my_corpus)

for d in diagnostics:
    print(f"[{d.severity.name}] {d.code} line {d.line}:{d.col} — {d.message}")

Expected output (no issues):

# (empty — no diagnostics means the template is clean)

Expected output (with issues):

[ERROR] E101 line 4:3 — Missing required frontmatter field: 'alias'
[WARNING] W201 line 9:1 — Tag 'draft' is not defined in the corpus

Prerequisites

• attune-rag is installed in your Python environment (pip install attune-rag)
• You have a corpus object that implements CorpusProtocol

Steps

1. Parse and validate frontmatter. Before linting a full template, check that its YAML frontmatter is well-formed:

   from attune_rag.editor import parse_frontmatter
   
   yaml_block = """
   alias: my-template
   tags: [draft, review]
   """
   
   data, issues = parse_frontmatter(yaml_block)
   for issue in issues:
       print(f"{issue.code}: {issue.message} (at {issue.path})")
   

   parse_frontmatter raises SchemaError if the YAML is malformed. issues is a list of FrontmatterIssue objects for schema violations — an empty list means the frontmatter is valid.

2. Lint a full template. Call lint_template with the file's text, its corpus-relative path, and the corpus. It returns a list of Diagnostic objects with 1-indexed line and column positions:

   from attune_rag.editor import lint_template
   
   diagnostics = lint_template(text, rel_path="templates/my_template.yaml", corpus=my_corpus)
   for d in diagnostics:
       print(d.to_dict())
   

3. Find references and plan a rename. Once the template is clean, you can locate all uses of a name and rename them across the corpus:

   from attune_rag.editor import find_references, plan_rename, apply_rename, ReferenceKind
   
   refs = find_references(my_corpus, name="old-alias", kind=ReferenceKind.ALIAS)
   print(f"Found {len(refs)} reference(s)")
   
   plan = plan_rename(my_corpus, old="old-alias", new="new-alias", kind=ReferenceKind.ALIAS)
   changed_paths = apply_rename(my_corpus, plan)
   print("Updated files:", changed_paths)
   

   apply_rename writes changes to disk and refreshes the corpus. It raises RenameCollisionError if new-alias already exists.

Next

Explore autocomplete for the template editor — call autocomplete_tags(corpus, prefix="dr") or autocomplete_aliases(corpus, prefix="my") to see prefix-matched suggestions returned as lists you can feed directly into an editor UI.