thedotmack/claude-mem / plugin/skills/agent-cost-report
agent-cost-report
VerifiedBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.
scripts/tests/test_labels.py
View on GitHubimport os
import unittest
import _paths # noqa: F401
from acr import labels
class Classify(unittest.TestCase):
def ev(self, prompts=(), text=()):
return dict(prompts=list(prompts), text=list(text))
def test_category_head_then_tail_then_default(self):
d = labels.classify(self.ev(["please fix the broken build"]), dict(project="p", user_prompt=None))
self.assertEqual((d["category"], d["matched"]["category_where"]), ("Bug fix", "head"))
d = labels.classify(self.ev([], ["benchmark results"]), dict(project="p", user_prompt=None))
self.assertEqual((d["category"], d["matched"]["category_where"]), ("Experiment", "tail"))
d = labels.classify(self.ev(), dict(project="p", user_prompt=None))
self.assertEqual((d["category"], d["matched"]["category_where"]), ("Investigation", "default"))
def test_failures_closed_set_and_rework_only_as_failure(self):
d = labels.classify(self.ev(["redo this, verifier found gaps", "x"] + ["same prompt prefix"] * 3), dict(project="p", user_prompt=None))
self.assertEqual(d["failure_signals"], ["Rework", "Looping"])
def test_active_minutes(self):
m = 60000
self.assertEqual(labels.active_minutes([0, 5 * m, 10 * m, 60 * m, 61 * m]), 11)
self.assertEqual(labels.active_minutes([]), 0)
class Review(unittest.TestCase):
def setUp(self):
self._prev_owner = os.environ.get("ACR_OWNER_NAME")
os.environ["ACR_OWNER_NAME"] = "Alex" # these cases are a configured owner, not a hardcoded one
def tearDown(self):
if self._prev_owner is None: os.environ.pop("ACR_OWNER_NAME", None)
else: os.environ["ACR_OWNER_NAME"] = self._prev_owner
def items(self):
return [dict(work_item_id="WI-1", category="Investigation", work_category="Investigation", failure_type="", failure_signals=[], label_source="keyword", status="completed"),
dict(work_item_id="WI-2", category="Investigation", work_category="Investigation", failure_type="", failure_signals=[], label_source="keyword", status="completed")]
def test_apply_sets_source_and_leaves_unreviewed(self):
li = self.items()
n = labels.apply_review(li, [dict(work_item_id="WI-1", category="Feature", failure_signals=["Looping", "Rework"], reviewed_by="alex", status="blocked")], "t")
self.assertEqual(n, 1); self.assertEqual((li[0]["label_source"], li[0]["failure_type"], li[0]["status"], li[0]["reviewed_at_pt"]), ("human", "Looping", "blocked", "t"))
self.assertEqual(li[1]["label_source"], "keyword")
labels.apply_review(li, [dict(work_item_id="WI-2", category="Feature", reviewed_by="model-x")], "t")
self.assertEqual(li[1]["label_source"], "llm")
def test_file_level_reviewed_by_default(self):
li = self.items()
n = labels.apply_review(li, dict(reviewed_by="Alex", items=[dict(work_item_id="WI-1", category="Feature")]), "t")
self.assertEqual((n, li[0]["label_source"], li[0]["reviewed_by"]), (1, "human", "Alex"))
self.assertEqual(labels.apply_review(li, dict(items=[dict(work_item_id="WI-2", category="Feature")]), "t"), 0)
def test_generated_review_template_inherits_file_level_reviewer(self):
li = self.items()
draft = labels.classify(dict(prompts=["build the feature"], text=[]), dict(project="p", user_prompt=None))
template = labels.review_entry(dict(li[0], session_ids=["session-1"], title="Feature work"), draft)
self.assertIsNone(template["reviewed_by"])
template["category"] = "Feature"
reviewed = dict(reviewed_by="Alex", items=[template])
self.assertEqual(labels.apply_review(li, reviewed, "t"), 1)
self.assertEqual((li[0]["reviewed_by"], li[0]["label_source"]), ("Alex", "human"))
self.assertIsNone(template["reviewed_by"])
def test_explicit_entry_reviewer_overrides_file_default(self):
li = self.items()
reviewed = dict(reviewed_by="Alex", items=[dict(work_item_id="WI-1", category="Feature", reviewed_by="model-x")])
self.assertEqual(labels.apply_review(li, reviewed, "t"), 1)
self.assertEqual((li[0]["reviewed_by"], li[0]["label_source"]), ("model-x", "llm"))
def test_empty_entry_reviewer_and_unconfirmed_category_remain_unreviewed(self):
li = self.items()
reviewed = dict(reviewed_by="Alex", items=[
dict(work_item_id="WI-1", category="Feature", reviewed_by=""),
dict(work_item_id="WI-2", category=None, reviewed_by=None),
])
self.assertEqual(labels.apply_review(li, reviewed, "t"), 0)
self.assertEqual([item["label_source"] for item in li], ["keyword", "keyword"])
def test_confirming_a_category_keeps_the_draft_failure_signals(self):
li = self.items(); li[0]["failure_signals"] = ["Recovery after miss"]; li[0]["failure_type"] = "Recovery after miss"
labels.apply_review(li, [dict(work_item_id="WI-1", category="Investigation", reviewed_by="Alex", failure_type=None, failure_signals=None)], "t")
self.assertEqual((li[0]["failure_signals"], li[0]["failure_type"], li[0]["label_source"]), (["Recovery after miss"], "Recovery after miss", "human"))
labels.apply_review(li, [dict(work_item_id="WI-1", category="Investigation", reviewed_by="Alex", failure_signals=[])], "t")
self.assertEqual((li[0]["failure_signals"], li[0]["failure_type"]), ([], ""))
def test_reviewed_title_replaces_the_draft_title(self):
li = self.items(); li[0]["title"] = li[1]["title"] = "first prompt text"
labels.apply_review(li, [dict(work_item_id="WI-1", category="Feature", reviewed_by="Alex", title="Rollup shipped"), dict(work_item_id="WI-2", category="Feature", reviewed_by="Alex", title="")], "t")
self.assertEqual((li[0]["title"], li[1]["title"]), ("Rollup shipped", "first prompt text"))
def test_unset_owner_does_not_treat_alex_as_a_human_review(self):
os.environ["ACR_OWNER_NAME"] = ""
li = self.items()
labels.apply_review(li, [dict(work_item_id="WI-1", category="Feature", reviewed_by="Alex")], "t")
self.assertEqual(li[0]["label_source"], "llm")
labels.apply_review(li, [dict(work_item_id="WI-2", category="Feature", reviewed_by="Alex", label_source="human")], "t")
self.assertEqual(li[1]["label_source"], "human")
def test_configured_owner_review_is_human_and_other_names_are_not(self):
os.environ["ACR_OWNER_NAME"] = "Sam"
li = self.items()
labels.apply_review(li, [dict(work_item_id="WI-1", category="Feature", reviewed_by="sam"),
dict(work_item_id="WI-2", category="Feature", reviewed_by="Alex")], "t")
self.assertEqual((li[0]["label_source"], li[1]["label_source"]), ("human", "llm"))
def test_full_owner_name_also_accepts_its_first_name(self):
os.environ["ACR_OWNER_NAME"] = "Sam Lee"
li = self.items()
labels.apply_review(li, [dict(work_item_id="WI-1", category="Feature", reviewed_by="sam"),
dict(work_item_id="WI-2", category="Feature", reviewed_by="Lee")], "t")
self.assertEqual((li[0]["label_source"], li[1]["label_source"]), ("human", "llm"))
signals = labels.classify(dict(prompts=["reset from sam: try again"], text=[]), dict(project="p", user_prompt=None))["failure_signals"]
self.assertIn("Rework", signals)
def test_reset_from_owner_is_rework_only_for_the_configured_name(self):
bare = dict(project="p", user_prompt=None)
def signals(text):
return labels.classify(dict(prompts=[text], text=[]), bare)["failure_signals"]
os.environ["ACR_OWNER_NAME"] = ""
self.assertNotIn("Rework", signals("reset from alex: try again"))
self.assertNotIn("Rework", signals("reset from scratch: wipe it"))
self.assertIn("Rework", signals("redo this"))
os.environ["ACR_OWNER_NAME"] = "Sam"
self.assertIn("Rework", signals("reset from Sam: try again"))
self.assertNotIn("Rework", signals("reset from alex: try again"))
self.assertNotIn("Rework", signals("reset from scratch: wipe it"))
os.environ["ACR_OWNER_NAME"] = "Alex"
self.assertIn("Rework", signals("reset from alex: try again"))
def test_rejects_unknown_values(self):
for bad in (dict(category="Rework"), dict(category="Feature", failure_type="Made up"), dict(category="Feature", label_source="magic"), dict(category="Feature", status="done")):
with self.assertRaises(labels.ReviewError):
labels.apply_review(self.items(), [dict(work_item_id="WI-1", reviewed_by="Alex", **bad)], "t")
if __name__ == "__main__":
unittest.main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| backup-setupthedotmack/claude-memSet up, check, or restore local claude-mem database snapshots. Use when the user says "set up backups", "back up my memories", "backup status", or "restore my database". | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |