oliverweissl/context-garden / skills/weeder
weeder
VerifiedMeasure and reduce a Skill's token footprint (always-loaded description + on-trigger SKILL.md body) without regressing routing accuracy or losing instructions. Use when a SKILL.md feels oversized, repeats rules, or has a vague description that might mis-route.
| File | Kind | Size |
|---|---|---|
| .gitignore | other | 32 B |
| SKILL.md | skill md | 3.1 KB |
| agents/weeder.md | markdown | 1.2 KB |
| bin/weederexecutable | script | 292 B |
| references/classification.md | markdown | 4.9 KB |
| references/llm-assist.md | markdown | 2.1 KB |
| references/routing-heuristic.md | markdown | 4.1 KB |
| scripts/cg_mode.py | script | 5.4 KB |
| scripts/we_apply_suggestion.py | script | 5.7 KB |
| scripts/we_audit.py | script | 5.6 KB |
| scripts/we_config.py | script | 2.5 KB |
| scripts/we_constraints.py | script | 5.1 KB |
| scripts/we_dupes.py | script | 6.1 KB |
| scripts/we_filler.py | script | 4.9 KB |
| scripts/we_optimize.py | script | 4.3 KB |
| scripts/we_parse.py | script | 5.0 KB |
| scripts/we_routing.py | script | 4.5 KB |
| scripts/we_suggest.py | script | 3.8 KB |
| scripts/we_text.py | script | 2.1 KB |
| scripts/we_tokens.py | script | 4.1 KB |
| scripts/weeder.pyexecutable | script | 12.0 KB |
| tests/fixtures/answer_lot.json | config | 875 B |
| tests/fixtures/filler_sentences.json | config | 2.6 KB |
| tests/fixtures/routing_examples.json | config | 489 B |
| tests/smoke_test.shexecutable | script | 14.3 KB |
scripts/we_filler.py
View on GitHub"""Filler-sentence detector for `audit`'s "likely unnecessary content".
Asymmetric by design: an imperative/constraint sentence is NEVER flagged
(deleting a real instruction is a silent regression; missed filler only costs
tokens), so unknown first words count as verbs. Rules: references/classification.md;
labelled fixture: tests/fixtures/filler_sentences.json.
"""
from __future__ import annotations
import re
from we_text import split_into_sentences
_CONSTRAINT_RE = re.compile(
r"\b(?:must|never|always|do\s+not|don['’]t|only|should|required|ensure|run|use|"
r"call|pass|set|avoid|prefer)\b",
re.I,
)
# Technically imperative, but conversational padding rather than an instruction.
_PLEASANTRY_OPENERS = re.compile(
r"^(?:please\s+)?(?:feel\s+free\b|note\s+that\b|note,|keep\s+in\s+mind\b|"
r"remember\s+that\b|bear\s+in\s+mind\b|thanks?\b|hope\b|enjoy\b|good\s+luck\b)",
re.I,
)
_NON_VERB_OPENERS = {
# pronouns / determiners
"this",
"that",
"these",
"those",
"it",
"it's",
"its",
"there",
"there's",
"here",
"here's",
"we",
"we're",
"you",
"you're",
"you'll",
"i",
"i'm",
"they",
"he",
"she",
"our",
"your",
"their",
"the",
"a",
"an",
"some",
"many",
"most",
"all",
"each",
"every",
"any",
"such",
"one",
"what",
"which",
"who",
"why",
"how",
"where",
# prepositions / conjunctions
"in",
"as",
"of",
"for",
"at",
"by",
"with",
"since",
"because",
"while",
"although",
"though",
"if",
"when",
"once",
"so",
"and",
"but",
"or",
"also",
"to",
"on",
"from",
"after",
"before",
"unlike",
"like",
"despite",
# hedges / discourse markers / pleasantries
"basically",
"generally",
"essentially",
"typically",
"usually",
"overall",
"ultimately",
"honestly",
"hopefully",
"obviously",
"clearly",
"of",
"great",
"happy",
"good",
"nice",
"perhaps",
"maybe",
"indeed",
"again",
"just",
"now",
"sometimes",
"often",
"however",
"otherwise",
"additionally",
"furthermore",
"moreover",
"anyway",
"welcome",
"congratulations",
"sure",
"ideally",
}
_FILLER_PATTERNS = [
# pleasantries
re.compile(r"\bfeel free to\b", re.I),
re.compile(r"\bhope (?:this|that|it) helps\b", re.I),
re.compile(r"\bhappy \w+ing\b", re.I),
re.compile(r"^(?:thanks|thank you)\b", re.I),
re.compile(r"\bgood luck\b", re.I),
re.compile(r"\bgreat question\b", re.I),
re.compile(r"\bwe hope\b", re.I),
re.compile(r"\bas an ai\b", re.I),
# hedges / throat-clearing
re.compile(r"\b(?:please )?note that\b", re.I),
re.compile(
r"\bit(?:'s| is) (?:important|worth|good) (?:to )?(?:note|noting|remember|mention)", re.I
),
re.compile(r"\bkeep in mind\b", re.I),
re.compile(r"\bbear in mind\b", re.I),
re.compile(r"\byou (?:might|may|could) (?:want|wish|find|like)\b", re.I),
re.compile(r"\bit (?:can|may|might|could) be (?:useful|helpful|handy|a good idea)\b", re.I),
re.compile(r"^(?:basically|generally|essentially|honestly|obviously|of course)\b", re.I),
re.compile(r"\bat the end of the day\b", re.I),
re.compile(r"\bin order to\b", re.I),
# rationale-only / self-description
re.compile(r"\bthis skill (?:is designed to|helps you|will|aims to|exists to)\b", re.I),
re.compile(
r"^(?:this|that|it)(?:'s| is| was) (?:very |really |quite |especially )?"
r"(?:useful|helpful|important|handy|nice|convenient|valuable|beneficial)\b",
re.I,
),
re.compile(r"^(?:this|that) (?:way|helps|makes|approach|means)\b", re.I),
re.compile(r"^the (?:reason|idea|goal|point) (?:for|behind|of|is)\b", re.I),
]
def _first_word(sentence: str) -> str:
m = re.match(r"[\s\"'(*_>]*([A-Za-z][A-Za-z'’]*)", sentence)
return m.group(1).lower().replace("’", "'") if m else ""
def starts_with_base_verb(sentence: str) -> bool:
s = re.sub(r"^(?:please|then|first|next|finally|now)[,\s]+", "", sentence.strip(), flags=re.I)
if _PLEASANTRY_OPENERS.match(s):
return False
w = _first_word(s)
if not w or w in _NON_VERB_OPENERS or "'" in w:
return False
if w.endswith(("ed", "ing", "ly")):
return False
if w.endswith("s") and not w.endswith(("ss", "us", "is")):
return False
return True
def is_instruction(sentence: str) -> bool:
return bool(_CONSTRAINT_RE.search(sentence)) or starts_with_base_verb(sentence)
def is_filler(sentence: str) -> bool:
s = sentence.strip()
if not s or is_instruction(s):
return False
return any(p.search(s) for p in _FILLER_PATTERNS)
def find_filler_sentences(text: str) -> list[str]:
return [s.strip() for s in split_into_sentences(text) if is_filler(s)]
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |