Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors
Rust
fix(docs): score a CRLF page like its LF twin in the eval examples
AkitaOnRails7a958c15f02c
diff --git a/docs/examples/auto-improve-eval/score_proposal.py b/docs/examples/auto-improve-eval/score_proposal.pyindex 46ef2bf3..26bf4d0a 100644--- a/docs/examples/auto-improve-eval/score_proposal.py+++ b/docs/examples/auto-improve-eval/score_proposal.py@@ -34,6 +34,9 @@ def has_text_beyond_headings(body: str) -> bool:def score_body(path: str, body: str) -> tuple[float, list[str]]:+ # A page written on Windows arrives with CRLF line endings; score it the+ # same as its LF twin instead of counting "\r" as text or as a list item.+ body = body.replace("\r\n", "\n").replace("\r", "\n")score = 0.0reasons: list[str] = []diff --git a/docs/examples/auto-improve-eval/score_proposal.sh b/docs/examples/auto-improve-eval/score_proposal.shindex aab34dfd..0891327b 100644--- a/docs/examples/auto-improve-eval/score_proposal.sh+++ b/docs/examples/auto-improve-eval/score_proposal.sh@@ -93,6 +93,7 @@ function score_body(path, body, lines, n, i, line, h1, h2, item, prose, placen = split(body, lines, "\n")for (i = 1; i <= n; i++) {line = lines[i]+ sub(/\r$/, "", line)if (line ~ /^#[ \t]+[^ \t]/) h1 = 1if (line ~ /^##[ \t]+[^ \t]/) h2 = 1if (line ~ /^[ \t]*([-*+]|[0-9]+[.)])[ \t]+[^ \t]/) item = 1diff --git a/docs/examples/auto-improve-eval/test_score_proposal.py b/docs/examples/auto-improve-eval/test_score_proposal.pyindex d5f827ec..d76d26ca 100644--- a/docs/examples/auto-improve-eval/test_score_proposal.py+++ b/docs/examples/auto-improve-eval/test_score_proposal.py@@ -93,6 +93,22 @@ class ScorerParityTest(unittest.TestCase):body = "Caminho C:\\n# x sem título\n"self.assert_verdict(proposal("_rules/a.md", body), False)+ def test_a_crlf_page_scores_as_its_lf_twin(self) -> None:+ # A page edited on Windows arrives with CRLF endings; the carriage+ # return must not count as text beyond a heading nor as a list item.+ cases = (+ ("_rules/x.md", "# Regra\n\nTexto.\n", True),+ ("_rules/x.md", "# Regra\n", False),+ ("procedures/x.md", "# P\n\n## S\n- passo\n", True),+ ("procedures/x.md", "# P\n\n## S\n- \n", False),+ )+ for path, body, passed in cases:+ lf = proposal(path, body)+ crlf = proposal(path, body.replace("\n", "\r\n"))+ self.assert_verdict(crlf, passed)+ for scorer in SCORERS:+ self.assertEqual(score(scorer, lf), score(scorer, crlf), scorer)+def test_a_non_json_input_fails_closed_in_the_shell_scorer(self) -> None:result = score("sh", b'{"path": "procedures/x", "after_body": "# a\n')self.assertIs(result["passed"], False)