All the Possibles

The report on designers.allthepossibles.com.

This is a sample, and it is a real one: the studio ran the report on its own Designers server on November 13, 2026 and published it unedited. Yours has the same sections for your list. Check your server free and the order form is on the result page.

designers.allthepossibles.com lists 141 tools that weigh about 40,000 tokens. Descriptions are 62% of that and input schemas are 33%.

The same report as files: the Markdown report for your repository and the annotations as JSON.

Where to cut first

The steps are ranked by how many tokens each one touches. Each number is the ceiling for its step, and the sections below give the detail.

  1. Cut the 105 descriptions that run past 600 characters. Together they weigh about 21,000 tokens.
  2. Shorten or remove the 5 repeated blocks of input schema. The copies after the first of each block weigh about 1,100 tokens.

Repeated blocks of input schema

5 blocks of 200 characters or more appear more than once in the list, word for word. The copies after the first of each block weigh about 1,100 tokens, which is 3% of the list.

One tool's schema cannot point at another tool's schema, so a block shared across tools has to be cut where it stands. Every character you trim from it comes off once per copy. When the shared text is an explanation, it can be said once in the server's instructions field, which a client receives at initialize, though not every client passes that field to the model. When several tools differ only in one value, one tool with a parameter replaces them.

  • About 130 tokens, 5 copies across 5 tools: site_inventory, site_heading_outline, site_meta_audit, site_consistency and site_performance_hints. The copies after the first cost about 530 tokens. About 70 tokens of each copy are description text. It first appears at inputSchema and begins: {"type":"object","properties":{"url":{"type":"string","description":"Home page or any page on the site. The crawl stays on this origin."},"paths":{"type":"array","items":{"type":"s
  • About 80 tokens, 4 copies across 4 tools: extract_design_tokens, audit_type_scale, audit_spacing and check_state_coverage. The copies after the first cost about 240 tokens. About 50 tokens of each copy are description text. It first appears at inputSchema and begins: {"type":"object","properties":{"css":{"type":"string","description":"CSS text to read (a stylesheet, or several concatenated). Free up to 100,000 characters; Pro up to 2,000,000."}
  • About 160 tokens, 2 copies across 2 tools: site_link_health and site_audit_summary. The copies after the first cost about 160 tokens. About 90 tokens of each copy are description text. It first appears at inputSchema and begins: {"type":"object","properties":{"url":{"type":"string","description":"Home page or any page on the site. The crawl stays on this origin."},"paths":{"type":"array","items":{"type":"s
  • About 90 tokens, 2 copies across 2 tools: check_motion and check_reflow. The copies after the first cost about 90 tokens. About 40 tokens of each copy are description text. It first appears at inputSchema and begins: {"type":"object","properties":{"html":{"type":"string","description":"Page HTML source. Optional if url is given."},"css":{"type":"string","description":"CSS text to analyze, with
  • About 50 tokens, 2 copies across 2 tools: alt_text_quality and repeated_label_check. The copies after the first cost about 50 tokens. It first appears at inputSchema and begins: {"type":"object","properties":{"url":{"type":"string","description":"A public page URL to fetch."},"html":{"type":"string","description":"HTML source to check instead of a URL."}},

Annotations

140 tools carry hints that nothing in the name contradicts, and the JSON below keeps them.

The proposals come from names and descriptions, not from calling anything, so read them before you paste. A tool that reads gets readOnlyHint: true. A tool that adds gets destructiveHint: false, and one that changes or removes gets destructiveHint: true, which is also what the spec assumes when the hint is missing.

  • write_design_changelog: It is marked readOnlyHint: true, but the name says it changes something that exists. If it does write, a client that trusts the hint runs it without asking the user, so read what it does against the hint.

Hints for 141 tools, keyed by tool name. Each object goes in that tool's annotations field.

{
 "alt_text_quality": {
  "readOnlyHint": true
 },
 "analyse_card_sort": {
  "readOnlyHint": true
 },
 "approx_print_colour": {
  "readOnlyHint": true
 },
 "audit_dark_patterns": {
  "readOnlyHint": true
 },
 "audit_email_accessibility": {
  "readOnlyHint": true
 },
 "audit_form": {
  "readOnlyHint": true
 },
 "audit_icon_set": {
  "readOnlyHint": true
 },
 "audit_motion_css": {
  "readOnlyHint": true
 },
 "audit_page_accessibility": {
  "readOnlyHint": true
 },
 "audit_site_accessibility": {
  "readOnlyHint": true
 },
 "audit_spacing": {
  "readOnlyHint": true
 },
 "audit_token_scales": {
  "readOnlyHint": true
 },
 "audit_type_scale": {
  "readOnlyHint": true
 },
 "brand_architecture_planner": {
  "readOnlyHint": true
 },
 "brand_voice_consistency": {
  "readOnlyHint": true
 },
 "build_grid": {
  "readOnlyHint": true
 },
 "build_heuristic_report": {
  "readOnlyHint": true
 },
 "build_spacing_scale": {
  "readOnlyHint": true
 },
 "button_label_consistency": {
  "readOnlyHint": true
 },
 "calc_fitts_hick": {
  "readOnlyHint": true
 },
 "calc_print_resolution": {
  "readOnlyHint": true
 },
 "calc_print_spec": {
  "readOnlyHint": true
 },
 "calculate_proportion": {
  "readOnlyHint": true
 },
 "chart_chooser": {
  "readOnlyHint": true
 },
 "chart_colour_check": {
  "readOnlyHint": true
 },
 "chart_lint": {
  "readOnlyHint": true
 },
 "check_cjk_line_breaking": {
  "readOnlyHint": true
 },
 "check_colour_mode": {
  "readOnlyHint": true
 },
 "check_contrast": {
  "readOnlyHint": true
 },
 "check_contrast_tokens": {
  "readOnlyHint": true
 },
 "check_design_drift": {
  "readOnlyHint": true
 },
 "check_email_accessibility": {
  "readOnlyHint": true
 },
 "check_email_dark_mode": {
  "readOnlyHint": true
 },
 "check_email_preheader": {
  "readOnlyHint": true
 },
 "check_email_subject": {
  "readOnlyHint": true
 },
 "check_font_coverage": {
  "readOnlyHint": true
 },
 "check_form": {
  "readOnlyHint": true
 },
 "check_handoff_spec": {
  "readOnlyHint": true
 },
 "check_input_modality": {
  "readOnlyHint": true
 },
 "check_keyboard_shortcuts": {
  "readOnlyHint": true
 },
 "check_motion": {
  "readOnlyHint": true
 },
 "check_name_fields": {
  "readOnlyHint": true
 },
 "check_naming_conventions": {
  "readOnlyHint": true
 },
 "check_proximity_alignment": {
  "readOnlyHint": true
 },
 "check_reading_order": {
  "readOnlyHint": true
 },
 "check_reflow": {
  "readOnlyHint": true
 },
 "check_rtl_readiness": {
  "readOnlyHint": true
 },
 "check_state_coverage": {
  "readOnlyHint": true
 },
 "check_svg_bloat": {
  "readOnlyHint": true
 },
 "check_touch_targets": {
  "readOnlyHint": true
 },
 "cognitive_load_check": {
  "readOnlyHint": true
 },
 "color_meaning_check": {
  "readOnlyHint": true
 },
 "colour_palette_brief_check": {
  "readOnlyHint": true
 },
 "colour_temperature_mix": {
  "readOnlyHint": true
 },
 "compare_token_sets": {
  "readOnlyHint": true
 },
 "content_hierarchy_check": {
  "readOnlyHint": true
 },
 "convert_paper_size": {
  "readOnlyHint": true
 },
 "css_contrast_pairs": {
  "readOnlyHint": true
 },
 "css_to_dtcg": {
  "readOnlyHint": true
 },
 "dark_mode_palette": {
  "readOnlyHint": true
 },
 "data_table_vs_chart": {
  "readOnlyHint": true
 },
 "design_easing": {
  "readOnlyHint": true
 },
 "detect_design_slop": {
  "readOnlyHint": true
 },
 "doctor": {
  "readOnlyHint": true
 },
 "draft_brand_checklist": {
  "readOnlyHint": true
 },
 "dtcg_to_css": {
  "readOnlyHint": true
 },
 "error_message_check": {
  "readOnlyHint": true
 },
 "error_state_writer_template": {
  "readOnlyHint": true
 },
 "extract_design_tokens": {
  "readOnlyHint": true
 },
 "generate_conformance_report": {
  "readOnlyHint": true
 },
 "get_project_context": {
  "readOnlyHint": true
 },
 "getting_started": {
  "readOnlyHint": true
 },
 "heuristic_worksheet": {
  "readOnlyHint": true
 },
 "inventory_components": {
  "readOnlyHint": true
 },
 "label_clarity_check": {
  "readOnlyHint": true
 },
 "lint_brand_voice": {
  "readOnlyHint": true
 },
 "lint_design_feedback": {
  "readOnlyHint": true
 },
 "lint_design_md": {
  "readOnlyHint": true
 },
 "lint_disclosure_copy": {
  "readOnlyHint": true
 },
 "lint_email_html": {
  "readOnlyHint": true
 },
 "lint_lottie_json": {
  "readOnlyHint": true
 },
 "lint_research_questions": {
  "readOnlyHint": true
 },
 "lint_token_names": {
  "readOnlyHint": true
 },
 "lint_tracking_plan": {
  "readOnlyHint": true
 },
 "localisation_readiness": {
  "readOnlyHint": true
 },
 "logo_contrast_variants": {
  "readOnlyHint": true
 },
 "logo_size_rules": {
  "readOnlyHint": true
 },
 "measure_leading": {
  "readOnlyHint": true
 },
 "modular_scale": {
  "readOnlyHint": true
 },
 "naming_check": {
  "readOnlyHint": true
 },
 "nav_depth_check": {
  "readOnlyHint": true
 },
 "notification_copy_check": {
  "readOnlyHint": true
 },
 "number_format_check": {
  "readOnlyHint": true
 },
 "page_weight_report": {
  "readOnlyHint": true
 },
 "palette_harmony": {
  "readOnlyHint": true
 },
 "plain_language_check": {
  "readOnlyHint": true
 },
 "plan_asset_exports": {
  "readOnlyHint": true
 },
 "plan_breakpoints": {
  "readOnlyHint": true
 },
 "plan_critique": {
  "readOnlyHint": true
 },
 "plan_durations": {
  "readOnlyHint": true
 },
 "plan_focus_order": {
  "readOnlyHint": true
 },
 "plan_imposition": {
  "readOnlyHint": true
 },
 "plan_responsive_images": {
  "readOnlyHint": true
 },
 "plan_stagger": {
  "readOnlyHint": true
 },
 "plan_text_expansion": {
  "readOnlyHint": true
 },
 "plan_usability_test": {
  "readOnlyHint": true
 },
 "print_file_checklist": {
  "readOnlyHint": true
 },
 "rendered_vs_static_notes": {
  "readOnlyHint": true
 },
 "repeated_label_check": {
  "readOnlyHint": true
 },
 "review_animation_principles": {
  "readOnlyHint": true
 },
 "review_date_number_formats": {
  "readOnlyHint": true
 },
 "review_locale_picker": {
  "readOnlyHint": true
 },
 "review_microcopy": {
  "readOnlyHint": true
 },
 "review_ui_copy": {
  "readOnlyHint": true
 },
 "save_project_context": {
  "readOnlyHint": false,
  "destructiveHint": true
 },
 "scan_pii": {
  "readOnlyHint": true
 },
 "score_design_system": {
  "readOnlyHint": true
 },
 "screen_reader_script": {
  "readOnlyHint": true
 },
 "site_audit_summary": {
  "readOnlyHint": true
 },
 "site_consistency": {
  "readOnlyHint": true
 },
 "site_heading_outline": {
  "readOnlyHint": true
 },
 "site_inventory": {
  "readOnlyHint": true
 },
 "site_link_health": {
  "readOnlyHint": true
 },
 "site_meta_audit": {
  "readOnlyHint": true
 },
 "site_page_diff": {
  "readOnlyHint": true
 },
 "site_performance_hints": {
  "readOnlyHint": true
 },
 "sitemap_to_outline": {
  "readOnlyHint": true
 },
 "stress_test_localisation": {
  "readOnlyHint": true
 },
 "tagline_review": {
  "readOnlyHint": true
 },
 "taxonomy_overlap": {
  "readOnlyHint": true
 },
 "tint_shade_scale": {
  "readOnlyHint": true
 },
 "touch_target_css": {
  "readOnlyHint": true
 },
 "tree_test_scorer": {
  "readOnlyHint": true
 },
 "trunk_test": {
  "readOnlyHint": true
 },
 "type_crimes_lint": {
  "readOnlyHint": true
 },
 "type_pairing_check": {
  "readOnlyHint": true
 },
 "url_structure_review": {
  "readOnlyHint": true
 },
 "usability_stats": {
  "readOnlyHint": true
 },
 "vertical_rhythm": {
  "readOnlyHint": true
 },
 "whoami": {
  "readOnlyHint": true
 },
 "write_design_changelog": {
  "readOnlyHint": false,
  "destructiveHint": true
 }
}

Descriptions that run long

105 descriptions run past 600 characters. A description has to tell the model when to pick the tool and what it returns. Worked examples and formatting rules can usually move to the parameter they belong to.

  • generate_conformance_report: about 390 tokens.
  • plan_breakpoints: about 320 tokens.
  • chart_lint: about 320 tokens.
  • type_crimes_lint: about 310 tokens.
  • type_pairing_check: about 300 tokens.
  • scan_pii: about 290 tokens.
  • check_proximity_alignment: about 290 tokens.
  • css_contrast_pairs: about 260 tokens.
  • score_design_system: about 250 tokens.
  • screen_reader_script: about 250 tokens.
  • number_format_check: about 250 tokens.
  • touch_target_css: about 250 tokens.
  • audit_form: about 250 tokens.
  • build_grid: about 250 tokens.
  • lint_email_html: about 240 tokens.
  • calculate_proportion: about 240 tokens.
  • plan_usability_test: about 240 tokens.
  • draft_brand_checklist: about 230 tokens.
  • build_spacing_scale: about 230 tokens.
  • chart_colour_check: about 230 tokens.
  • dtcg_to_css: about 230 tokens.
  • stress_test_localisation: about 220 tokens.
  • measure_leading: about 220 tokens.
  • modular_scale: about 220 tokens.
  • compare_token_sets: about 220 tokens.
  • inventory_components: about 220 tokens.
  • audit_dark_patterns: about 220 tokens.
  • calc_fitts_hick: about 220 tokens.
  • check_form: about 220 tokens.
  • audit_motion_css: about 210 tokens.
  • repeated_label_check: about 210 tokens.
  • calc_print_spec: about 210 tokens.
  • error_message_check: about 210 tokens.
  • localisation_readiness: about 210 tokens.
  • cognitive_load_check: about 200 tokens.
  • check_rtl_readiness: about 200 tokens.
  • site_link_health: about 200 tokens.
  • color_meaning_check: about 200 tokens.
  • analyse_card_sort: about 200 tokens.
  • page_weight_report: about 200 tokens.

Parameters with no description

16 parameters across 4 tools have no description. The model fills those from the parameter name alone. A name like id or q is where wrong arguments come from, and a name that explains itself needs nothing.

  • data_table_vs_chart: 9 of 9.
  • plan_responsive_images: 4 of 9.
  • plan_usability_test: 2 of 6.
  • number_format_check: 1 of 5.

Every tool by weight

Tokens for the name, description and input schema together, then the description and the input schema on their own.

ToolTokensDescriptionInput schemaParameters
plan_breakpoints59831726610
generate_conformance_report5753861727
modular_scale4822222479
calc_fitts_hick4702182388
build_grid4492451918
check_proximity_alignment4492881444
calc_print_spec43321020910
score_design_system4202541504
chart_chooser4121972018
number_format_check4072501425
calculate_proportion4022411465
usability_stats39910228312
css_contrast_pairs3972561255
scan_pii393294862
type_crimes_lint393309692
stress_test_localisation3892231493
chart_lint381316521
plan_responsive_images3811941719
type_pairing_check380301631
plan_usability_test3782391236
site_link_health3772021605
draft_brand_checklist3762331274
measure_leading3752231375
compare_token_sets3722221354
localisation_readiness3702061494
build_spacing_scale3692301245
logo_size_rules3681611936
calc_print_resolution3581981447
error_state_writer_template3571711698
site_audit_summary3531771605
audit_dark_patterns3502201143
touch_target_css347249834
inventory_components3442221073
chart_colour_check342228983
lint_brand_voice3411901362
site_performance_hints3381901324
dtcg_to_css333225943
error_message_check3322101071
audit_form330249683
nav_depth_check3291761384
plan_imposition3211951124
screen_reader_script317253492
vertical_rhythm3121921034
brand_voice_consistency3111751203
data_table_vs_chart3091721219
color_meaning_check308201911
content_hierarchy_check302187983
naming_check3011821063
colour_temperature_mix300196881
site_consistency2991521324
site_inventory2981521324
dark_mode_palette297196861
audit_motion_css297212713
check_motion296195873
palette_harmony2951791022
analyse_card_sort295201802
lint_disclosure_copy294179992
audit_token_scales2941561233
logo_contrast_variants2941421363
tree_test_scorer292182952
check_reflow291190873
plan_durations2911341435
site_meta_audit2901431324
extract_design_tokens289193792
review_date_number_formats289186873
taxonomy_overlap2881511223
check_contrast285971746
design_easing2841601105
lint_email_html284242271
brand_architecture_planner282191742
repeated_label_check281212522
site_heading_outline2811331324
check_state_coverage275180792
cognitive_load_check275204551
check_design_drift274189702
check_font_coverage272187702
check_form271215432
plan_text_expansion271175813
write_design_changelog270169854
audit_type_scale266172792
button_label_consistency265162862
notification_copy_check265163862
plan_stagger2641271235
alt_text_quality260192522
check_colour_mode260172733
trunk_test259198472
check_rtl_readiness259204401
check_handoff_spec258195491
check_keyboard_shortcuts256164752
build_heuristic_report255811583
check_email_preheader255171683
page_weight_report253200371
check_input_modality249193401
audit_spacing248155792
css_to_dtcg248192431
colour_palette_brief_check248148832
review_microcopy248178561
convert_paper_size247150824
review_animation_principles244156712
check_cjk_line_breaking244185421
audit_email_accessibility243186391
plan_asset_exports240178471
tint_shade_scale238169542
print_file_checklist237132903
tagline_review235130913
review_locale_picker235176441
check_svg_bloat235187341
plan_focus_order234180391
plain_language_check233128893
sitemap_to_outline231136802
check_email_accessibility230186271
check_email_dark_mode229186271
url_structure_review227164471
audit_icon_set223169401
check_name_fields222166401
check_naming_conventions220161421
check_touch_targets219831212
check_reading_order216161401
audit_page_accessibility206115743
label_clarity_check206143481
doctor204116763
lint_tracking_plan20195913
lint_lottie_json198156281
plan_critique195571255
approx_print_colour192148291
rendered_vs_static_notes191126461
heuristic_worksheet19082934
check_email_subject175130291
site_page_diff17196612
review_ui_copy16189582
detect_design_slop16092532
check_contrast_tokens14397301
lint_research_questions13467512
audit_site_accessibility11657422
lint_design_md11269291
lint_design_feedback11063321
save_project_context11047482
lint_token_names10964311
getting_started10856381
get_project_context5716261
whoami492880

How this was measured

The report was built from initialize and tools/list, and no tool was called. Tokens are counted at four characters each over the name, description and input schema of every tool, which is the part clients pass to the model. The address read was https://designers.allthepossibles.com/mcp, on 2026-11-13.

The report on your own server costs $49 and is ready a few seconds after you pay. Start with the free check.