agent-parameter-categorizer

An API-parameter sorting guide that assigns each parameter to a fixed tier using a decision tree. An API is a defined way for software systems to communicate.

In plain words
What is it for?
It helps categorize parameters as required inputs, filters, range settings, output controls, or standard options, then order them, explain the decisions, and verify consistent results.
Why use it?
It removes inconsistent manual ordering and makes the same kind of parameter appear in the same place across tools.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/yaleh/meta-cc/agent-parameter-categorizer
Clone the repo
git clone --depth 1 https://github.com/yaleh/meta-cc
Per session 32 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,402 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00032 $0.02402
Opus 5 $0.00016 $0.01201
Sonnet 5 $0.00006 $0.00480
Haiku 4.5 $0.00003 $0.00240

Measured 2d ago against content hash 71318fa2d481, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-parameter-categorizer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

_experiments/bootstrap-006-api-design/agents/agent-parameter-categorizer.md · 294 lines

How it starts

The opening of the file, as written. The whole thing — 294 lines — stays where its author put it; the contents beside it link to each section on GitHub.

λ(api, parameters) → categorization | ∀param ∈ parameters:

categorize :: (API, Parameters) → Categorization categorize(A, P) = read(A) → apply_tree(P) → group_by_tier() → order() → add_comments() → verify_determinism() → process_all() → report()

apply_decision_tree :: Parameter → Tier apply_decision_tree(p) = { tier: match p with | _ where p.required = true → 1, | _ where filters_results(p) → 2, | _ where defines_range(p) → 3, | _ where controls_output(p) → 4, | _ where is_standard(p) → 5, | _ → unclassified,

return {parameter: p.name, tier: tier, reason: explain_decision(p, tier)} }

filters_results :: Parameter → Bool filters_results(p) = (p.description ~ "filter|narrow|select|match") ∧ ¬(p.description ~ "limit|offset|count|size")

defines_range :: Parameter → Bool defines_range(p) = (p.name ~ "min_|max_|threshold|window|start_|end_") ∨ (p.description ~ "bound|range|minimum|maximum")

controls_output :: Parameter → Bool controls_output(p) = p.name ∈ {"limit", "offset", "output_format", "max_results"} ∨ (p.description ~ "output size|result count|format")

is_standard :: Parameter → Bool is_standard(p) = p.name ∈ {"scope", "jq_filter", "stats_only", "stats_first", "inline_threshold_bytes"}

categorize_all :: Parameters → Categorized_Parameters categorize_all(P) = { categorized: { tier_1: [p | p ∈ P ∧ apply_decision_tree(p).tier = 1], tier_2: [p | p ∈ P ∧ apply_decision_tree(p).tier = 2], tier_3: [p | p ∈ P ∧ apply_decision_tree(p).tier = 3], tier_4: [p | p ∈ P ∧ apply_decision_tree(p).tier = 4], tier_5: [p | p ∈ P ∧ apply_decision_tree(p).tier = 5], unclassified: [p | p ∈ P ∧ apply_decision_tree(p).tier = unclassified] },

return categorized }

order_by_tier :: Categorized_Parameters → Ordered_Parameters order_by_tier(C) = { ordered: flatten([ C.tier_1, C.tier_2, C.tier_3, C.tier_4, C.tier_5 ]),

return { parameters: ordered, sequence: [p.name | p ∈ ordered] } }

add_tier_comments :: Ordered_Parameters → Commented_Parameters add_tier_comments(O) = { with_comments: [],

∀tier ∈ [1, 2, 3, 4, 5] → params_in_tier: [p | p ∈ O.parameters ∧ p.tier = tier],

if |params_in_tier| > 0 then
  comment: generate_tier_comment(tier),
  with_comments += [comment] + params_in_tier,

return with_comments }

generate_tier_comment :: Tier → Comment generate_tier_comment(tier) = { comments: { 1: "// Tier 1: Required Parameters", 2: "// Tier 2: Filtering", 3: "// Tier 3: Range Parameters", 4: "// Tier 4: Output Control", 5: "// Tier 5: Standard Parameters (added automatically)" },

return comments[tier] }

verify_determinism :: Categorization → Determinism_Metrics verify_determinism(C) = { total: |C.all_parameters|, categorized: |C.tier_1| + |C.tier_2| + |C.tier_3| + |C.tier_4| + |C.tier_5|, ambiguous: |C.unclassified|, determinism_rate: categorized / total,

checks: { single_tier: ∀p ∈ C.all_parameters → count(t | p ∈ C[t]) = 1, consistent: ∀p ∈ C.all_parameters → apply_decision_tree(p).tier ≠ unclassified, ordered: sequence_matches_tier_order(C.ordered) },

return { total_parameters: total, categorized: categorized, ambiguous: ambiguous, determinism_rate: determinism_rate, all_checks_passed: checks.single_tier ∧ checks.consistent ∧ checks.ordered } }

process_all_tools :: Tools → Processing_Results process_all_tools(T) = { results: [],

∀tool ∈ T → parameters: extract_parameters(tool), categorization: categorize_all(parameters), ordering: order_by_tier(categorization), commented: add_tier_comments(ordering),

compliance_before: calculate_compliance(tool.current_order, ordering.sequence),

apply_reordering(tool, commented),

compliance_after: calculate_compliance(tool.new_order, ordering.sequence),

run_tests(tool),

results += {
  tool: tool.name,
  parameters_categorized: |parameters|,
  ambiguous_cases: |categorization.unclassified|,
  compliance_before: compliance_before,
  compliance_after: compliance_after,
  status: if compliance_after = 1.0 then "COMPLETED" else "NEEDS_REVIEW"
},

Read the full file on GitHub · 294 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 294 lines · 32 tokens per session scan A 71318fa2d481

Subscribe to this mod's changes

agent-parameter-categorizer is an agent published in the GitHub repository yaleh/meta-cc (21 stars, last pushed 11d ago), licensed MIT. It adds 32 tokens to every session and 2,402 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.