agentleFS
peterbamuhigirepeterbamuhigire/proposal-skills/skills/ai-agent-proposals/ai-agent-pricing-and-packaging-proposal/SKILL.md

ai-agent-pricing-and-packaging-proposal

Skill11 starsChanged 16 days agoMITView on GitHub

Use when selecting and presenting per-resolution, per-outcome, per-step, per-agent, hybrid, or success-based pricing for an AI agent; use commercial-layer skills for contract exhibits.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write the intervention-credit clause — if intervention rate exceeds the agreed ceiling, the buyer receives a credit, calibrated to the …

    line 43
  2. Distinctive:

    Write the vendor-cost-pass-through clause — provider model-price index, FX index, annual cap, margin floor.

    line 44
  3. Distinctive:

    Write the fair-use clause specific to agents — per-tenant action ceiling per day, anomalous-action ceiling, scope-creep trigger (a new …

    line 45
  4. Distinctive:

    Write the abort and refund clause — what triggers a refund (irreversible-action incident at the agency's fault.

    line 46
  5. Distinctive:

    Write the autonomy-ramp clause — pricing assumes a stated autonomy ramp.

    line 47
  6. Distinctive:

    No intervention credit (buyer carries all the supervisor cost).

    line 132
  7. Distinctive:

    Verify acceptance conditions, permission boundaries, direct references, and anti-slop controls.

    line 223

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AI-Agent Pricing and Packaging Proposal

Acknowledgement: Shared by Peter Bamuhigire, techguypeter.com, +256 784 464178.

<!-- dual-compat-start -->

Use When

  • The AI-agent product or service is being priced for the first time, repriced, or repackaged.
  • The buyer expects a defensible pricing structure that survives CFO and procurement scrutiny.
  • The proposal must address per-resolution, per-outcome, per-step, per-agent, hybrid, or success-based pricing.
  • Vendor model-call cost behaviour is volatile and the agency needs cost-pass-through with a margin floor.
  • The competitive set includes vendors who quote "per resolution" without defining what a resolution is, or quote a per-seat licence as if the agent did not consume tokens.

Do Not Use When

  • The pricing is for a non-agent SaaS or AI feature (use saas-pricing-and-packaging-proposal or ai-on-saas-pricing-and-packaging-proposal).
  • The engagement is fee-for-service AI consulting with no recurring agentic operations.

Domain Inputs

  • The Agent Value Stack and Agent Cost Stack from ai-agent-business-case-and-roi.
  • The buyer's commercial instinct (subscription, usage, outcome, hybrid).
  • The competitive pricing reference points (named competitors, prices, definitions of "resolution").
  • Model-call cost per task at P50 / P90 / P99 per use case.
  • Buyer's tolerance for variable bills and for outcome-based risk-sharing.
  • Buyer's tier strategy (good / better / best, or persona-based, or sovereign / on-prem).
  • The intervention-rate target curve from the pilot.
  • The reversibility classification of each action class.

Domain Method

  1. Decide the agent pricing pattern from the six patterns (below).
  2. Map the pattern to the buyer's tolerance, the cost stack, and the intervention-rate target. Per-resolution looks attractive but is dangerous if intervention rate stays high. Per-step is conservative for the agency but invites scope-creep games.
  3. Decide what counts (definition of a resolution, an outcome, a step, an agent). Procurement will exploit any ambiguity.
  4. Write the intervention-credit clause — if intervention rate exceeds the agreed ceiling, the buyer receives a credit, calibrated to the intervention overhead the buyer carried.
  5. Write the vendor-cost-pass-through clause — provider model-price index, FX index, annual cap, margin floor. Agents fan out model calls; a model-price increase that is invisible in a copilot can be catastrophic in an agent.
  6. Write the fair-use clause specific to agents — per-tenant action ceiling per day, anomalous-action ceiling, scope-creep trigger (a new action class added by the buyer requires a change order).
  7. Write the abort and refund clause — what triggers a refund (irreversible-action incident at the agency's fault; intervention rate beyond ceiling for 60 consecutive days; regulator action against the agentic system).
  8. Write the autonomy-ramp clause — pricing assumes a stated autonomy ramp; if the buyer demands a higher autonomy level than the methodology supports, the buyer accepts the resulting risk and the price changes.
  9. Show worked examples — small tenant, medium tenant, heavy tenant — for each pattern.
  10. Output the Agent Pricing Proposal subsection of the financial proposal.

The Six Agent Pricing Patterns

Pattern A — Per-Resolution (Outcome Unit)

  • The agency charges $X per resolved ticket / triaged claim / reconciled batch / drafted document.
  • "Resolution" is defined narrowly (the agent did the work; the human did not have to redo it; the outcome held for N days).
  • Best for: high-volume, well-defined outcome workflows (support, claims acknowledgement, low-stakes operations).
  • Risk to agency: intervention rate exceeds plan and the price-per-resolution does not cover model and supervision cost.
  • Mitigation: floor price, intervention-credit ceiling, model-cost-pass-through clause.

Pattern B — Per-Outcome (Business Result Unit)

  • The agency charges per business outcome, not per agent action. Example: per recovered debt, per renewed policy, per closed bug.
  • Best for: workflows where the buyer can attribute outcome cleanly to the agent.
  • Risk to agency: attribution disputes; outcome timing exceeds agency control.
  • Mitigation: outcome definition with cooling-off period and dispute clause; minimum-fee floor.

Pattern C — Per-Step (Action Unit)

  • The agency charges per agent action (tool call, system write, message sent), or per task step.
  • Best for: workflows where the buyer wants visibility and refuses to pay for outcomes the agent does not control end-to-end.
  • Risk to agency: incentive for the agent to over-call tools.
  • Mitigation: step budget per task; anomaly alerting; quarterly review.

Pattern D — Per-Agent (Seat / Persona Unit)

  • The agency charges per agent persona (e.g. "1 support agent — 24x7 — handling up to N tickets / month").
  • Best for: buyers who think in headcount terms and want a clean comparison ("this agent costs the equivalent of 0.3 FTE").
  • Risk to agency: heavy users overrun the included volume; light users under-utilise.
  • Mitigation: included-volume cap and overage rate; volume-floor for the buyer.

Pattern E — Hybrid (Base + Usage)

  • A base fee covers platform, oversight queue, eval and red-team, plus a per-resolution / per-step rate above an included allowance.
  • Best for: enterprise buyers who want budget predictability with a usage tail.
  • Risk to agency: complex bills, hard to explain.
  • Mitigation: clear allowance, overage rate, monthly statement, quarterly review.

Pattern F — Success-Based (Outcome-Linked Bonus)

  • A base fee at a lower rate plus a bonus on achieving an agreed outcome threshold (e.g. 80 % autonomy at month 6 → bonus; intervention rate below 10 % → bonus).
  • Best for: trust-building engagements where the agency has confidence in the methodology and the buyer wants skin-in-the-game.
  • Risk to agency: outcome shortfall destroys margin.
  • Mitigation: bonus is genuinely a bonus, not buried as risk; base fee covers cost; cap on bonus.

Worked Pattern Map by Use Case

Use caseRecommended patternWhy
Customer support resolutionA or EVolume is high; resolution is defined; both sides care about per-ticket economics
Insurance claims triageC or DOutcome attribution is messy; per-step or per-agent is cleaner
Financial reconciliationB or DOutcome is clean (reconciled batch); per-agent works for in-house ops
Legal drafting / reviewC or DOutcome is the human's; per-step or per-agent fits
Citizen-service (public-sector)DOutcome-based pricing is politically inappropriate; per-agent gives transparency
Healthcare adminDOutcome attribution is sensitive; per-agent gives clarity
Coding / ops triageC or EStep count and base + usage suit engineering ops

Clauses the Procurement Team Will Read First

  • Definition of unit — every pricing unit (resolution, outcome, step, agent) defined in one paragraph with examples and counter-examples.
  • Intervention credit — if intervention rate exceeds X % for the month, the buyer receives a Y % credit on units billed.
  • Vendor cost pass-through — model-call cost indexed to provider price; annual cap (e.g. CPI + 3 % or model-price-index + N %); margin floor; notice before increase.
  • Fair-use — per-tenant action ceiling per day; anomalous-action ceiling (e.g. agent attempts more than 3× P95 daily action volume); scope-creep trigger (new action class requires change order).
  • Abort and refund — irreversible-action incident at agency fault → defined refund; intervention rate beyond ceiling for 60 days → buyer may exit with pro-rata refund; regulator action → joint review.
  • Autonomy ramp — pricing assumes ramp curve; demand for faster ramp invokes a re-price.
  • FX — model costs in USD; revenue in local currency; FX index or annual reset.

Quality Standards

  • The pricing pattern is named and justified by the cost stack and the buyer's tolerance.
  • Cost-of-model-calls at P99 per task is below the priced unit, with margin, in every worked example.
  • Definitions of "resolution / outcome / step / agent" are explicit.
  • Intervention credit and abort-and-refund clauses are present.
  • Vendor cost pass-through has an index, cap, and margin floor.
  • Fair-use is defined numerically.
  • FX is named.
  • The pricing protects both sides against model-price volatility and against intervention overshoot.

Domain Risks

  • Per-resolution pricing with "resolution" undefined.
  • "Outcome-based pricing" with no attribution clause.
  • Per-step pricing with no step budget.
  • Per-agent pricing with no included-volume cap.
  • Hybrid pricing with the allowance hidden in an appendix.
  • Success bonus written as risk dressed up as bonus.
  • Model-cost pass-through as "subject to change".
  • No intervention credit (buyer carries all the supervisor cost).
  • Scope-creep ignored — buyer adds use cases at no extra price and the agency drowns.

Domain Outputs

  • Agent Pricing Proposal subsection of the financial proposal.
  • Agent Pricing Pattern Decision Memo (win-room file).
  • Unit Definitions (resolution / outcome / step / agent) one-paragraph each.
  • Intervention Credit Clause.
  • Vendor Cost Pass-Through Clause.
  • Fair-Use Clause.
  • Abort-and-Refund Clause.
  • Autonomy-Ramp Clause.
  • Worked examples for small / medium / heavy tenants.

Packaging-Pattern Decision Matrix

The pricing pattern is paired with a packaging shape — Included in Pro, Add-on, or Standalone — that constrains the SLA class achievable, the credit cap, the renewal posture, and the cost recovery shape. Use ai-agent-commercial-packaging to choose the shape before finalising the pattern. The reference matrix ai-agent-packaging-pattern-decision-matrix shows pattern × shape combinations with worked examples.

ShapeRecommended patternsDefault SLA classCredit cap
Included in ProD, EBronze / Silver≤ 25 %
Add-onA, C, D, ESilver / Gold25–50 %
StandaloneB, F, E (high-base)Gold / Platinum50–100 %

SLA-Tier Alignment with Pricing Tier

Every pricing pattern attaches to an SLA Class. The SLA Class drives the metric thresholds, the credit schedule, the kill-switch and audit-log SLAs, and the credit cap. The pricing exhibit and the SLA exhibit must reference the same numbers, same definitions, and same evidence trail. Load ai-agent-sla-and-credit-schedule after choosing the pattern.

Commercial Layer Cross-Links

The pricing pattern is one of nine commercial artefacts the proposal carries. The others are produced by the dedicated commercial-layer skills:

  • ai-agent-sla-and-credit-schedule — SLA class and credit schedule.
  • ai-agent-commercial-packaging — packaging shape decision.
  • ai-agent-contract-language-pack — drop-in exhibit assembly.
  • ai-agent-success-fee-and-outcome-pricing — outcome and success-fee structures (Patterns B, F, E with success component).
  • ai-agent-intervention-credit-and-abort-refund — intervention-credit formula and abort-refund triggers.
  • ai-agent-msa-and-sla-addendum-templates — MSA / SLA addendum.
  • ai-agent-procurement-objections-on-commercials — commercial objection handling.
  • ai-agent-renewal-and-true-up — renewal, true-up, autonomy price-step.
  • ai-agent-pricing-exhibit-template — drop-in pricing exhibit.

Anti-Patterns

  • Inventing a metric, credential, constraint, or buyer position. Fix: cite the supplied source or mark the item as an assumption requiring confirmation.
  • Treating an unavailable check as passed. Fix: mark it not assessed and state the evidence needed to resume.
  • Advancing autonomy without a named gate owner. Fix: require observable evidence, accountable acceptance, and a rollback path.
  • Reusing another sector or use case without reassessment. Fix: retest affected parties, action scope, reversibility, and jurisdiction.
  • Writing acceptance as “satisfactory” or “appropriate”. Fix: define an observable measure, threshold, evidence record, and decision owner.

Inputs

ArtefactSource/providerRequired?Missing-input behaviour
scope, action volumes, cost stack, outcome definition, intervention forecast, and buyer buying modelBuyer evidence, ToR, approved discovery record, system owner, or measured operating dataYesStop the affected decision; list the missing source and return only a qualified outline or assumption register.

Outputs

ArtefactConsumerAcceptance condition
Agent pricing and packaging exhibitBuyer CFO, procurement, and proposal evaluatorScope, assumptions, exclusions, owners, decision logic, and observable acceptance tests are explicit and traceable to supplied evidence.

Evidence Produced

EvidenceConsumerAcceptance condition
agent pricing and packaging exhibitBuyer CFO, procurement, and proposal evaluatorEvery load-bearing claim traces to supplied evidence; assumptions, owners, gates, exclusions, and observable acceptance conditions are explicit.

Capability Contract

Default to read-only for discovery, analysis, review, and planning. Minimum capability is access to the supplied artefacts and permission to calculate or inspect evidence. Edit only the requested proposal working copy. Do not change production systems, contact affected parties, publish, spend, certify compliance, or approve autonomous action without explicit authority from the accountable owner.

Degraded Mode

If files, interviews, telemetry, specialist review, network access, or calculation tools are unavailable, produce the narrowest useful qualified result. Mark each unavailable check as not assessed, separate facts from assumptions, lower confidence, and state the evidence needed to resume. An unassessed gate is never a pass.

Decision Rules

ChoiceActionFailure or risk avoided
Choose pricing patternMatch the billable unit to measurable buyer value, cost drivers, attribution, and auditability.Pricing that rewards attempts, hides cost, or cannot be reconciled.
Required evidence, authority, or accountable owner is missingStop the affected recommendation or commitment and record the gap.Invented evidence or unauthorised autonomy.
Gate evidence is complete and acceptedAdvance only within the approved scope and retain the evidence trace.Scope drift and irreproducible approval.

Workflow

  1. Confirm the consumer, authority, neighbouring-skill route, and required inputs; stop when a mandatory source or accountable owner is missing.
  2. Inspect the evidence and record facts, assumptions, conflicts, and unavailable checks; stop on a failed safety, finance, regulatory, or acceptance gate.
  3. Apply the domain method and decision rules within the qualified scope, retaining an evidence trace.
  4. Draft the contracted output and reconcile it with methodology, work plan, staffing, pricing, risk, and governance; recover by revising the affected scope or control and rerunning the failed gate.
  5. Verify acceptance conditions, permission boundaries, direct references, and anti-slop controls; block release until failed checks are corrected.

Worked Example

For a resolution agent, bill qualified resolutions with an intervention credit and minimum commitment; separate upstream model-cost pass-through and define disputed-resolution handling.

<!-- dual-compat-end -->

References

  • ai-agent-pricing-models-reference — pattern library with worked examples and clauses.
  • ai-agent-pricing-exhibit-template — drop-in pricing exhibit.
  • ai-agent-packaging-pattern-decision-matrix — packaging decision matrix.
  • ai-agent-sla-class-table — SLA class thresholds and credit schedule.
  • ai-on-saas-pricing-models-reference — AI-on-SaaS pricing patterns.
  • ai-on-saas-pricing-and-packaging-proposal — AI-on-SaaS pricing when the agent lives inside a SaaS product.
  • ai-agent-business-case-and-roi — cost stack input.
  • ai-agent-sla-and-credit-schedule — SLA class and credit schedule.
  • ai-agent-commercial-packaging — packaging shape.
  • 10-financial-proposal — financial proposal placement.
  • premium-pricing-and-value-defense — premium-fee defence.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.