agentleFS
peterbamuhigirepeterbamuhigire/business-plan-skills/skills/meta-finance/meta-agent-valuation-adjustments/SKILL.md

meta-agent-valuation-adjustments

Skill4 starsChanged 19 days agoMITView on GitHub

Use when an agent-product plan is being valued. Use financial projections for model construction.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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:

    Multi-agent governance discount: if multi-agent product without branch / loop caps + kill-switch + governance evidence = -20-40%

    line 71
  2. Distinctive:

    All three layers (SaaS / AI / Agent) adjustments visible and computed transparently

    line 89

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
<!-- dual-compat-start -->

Meta — Agent Valuation Adjustments Skill

Overview

SaaS valuation (handled by meta-valuation + saas-valuation-and-fundraising-strategy) uses Rule-of-40-adjusted ARR multiples + cohort discipline. AI valuation (handled by meta-ai-valuation-adjustments) layers premium / discount for AI defensibility. Agent valuation must add a separate adjustment because agent businesses span a much wider range — from "genuinely defensible vertical agent" earning a substantial premium to "GPT prompt wrapper" earning a wrapper discount of 40-70% on the headline ARR multiple.

The dispersion of agent-business multiples in 2025-2026 is wider than any other AI category. Investors are explicit about premium and discount; the model must be too.

Use When

  • An agent-product plan is being valued
  • A funding round is being planned and pre-money expectation set
  • An exit comparable analysis is being run
  • A board pack includes valuation update
  • The plan must declare its expected multiple range with reasoning
  • Cross-loaded with meta-valuation and meta-ai-valuation-adjustments

Do Not Use When

  • The product is not an agent — use SaaS / AI valuation alone
  • The business is too early for valuation discipline (pre-revenue with no operating evidence) — use directional range only
  • Route to 10-financial-projections instead when the task is to construct the underlying model.

Required Inputs

InputSource / providerRequired?If absent
Valuation Adjustments brief and decision audienceClient, plan owner, or approved project filesYesStop before making a recommendation; state the missing decision context.
Claims, assumptions, and supporting evidenceSource register, model, research notes, interviews, or operating recordsYesSeparate known facts from assumptions and return a qualified gap list.
Authority and delivery constraintsRequesting owner and repository instructionsYesRemain read-only and produce a draft or review only.
Current accounting, tax, valuation, or pricing basisFinance owner, accounting records, signed contracts, and current authoritative sourcesConditionalMark the treatment unresolved and require qualified professional review.
  • SaaS valuation base (from meta-valuation and saas-valuation-and-fundraising-strategy)
  • AI valuation adjustment (from meta-ai-valuation-adjustments)
  • Agent moat-vs-wrapper score (from saas-agent-moat-and-wrapper-risk)
  • Agent unit economics (cost-per-resolved-task, agent GM)
  • Intervention rate trajectory
  • Task success rate trajectory
  • Irreversibility-incident log
  • Regulator engagement evidence
  • Comparable transactions in agent space (recent rounds, recent exits)
  • Foundation-model-provider trajectory in your category

Workflow

  1. Compute the SaaS base multiple — Rule-of-40-adjusted ARR multiple from meta-valuation.
  2. Apply the AI adjustment — premium or discount from meta-ai-valuation-adjustments.
  3. Apply the agent adjustment per references/saas-agent-valuation-adjustments.md:
    • Moat-vs-wrapper score (0-24 from saas-agent-moat-and-wrapper-risk):
      • 0-8 (wrapper): apply 40-70% discount on AI-adjusted multiple
      • 9-14 (real but incomplete): apply 0-20% discount or 0% premium depending on trajectory
      • 15-19 (strong moat): apply 20-50% premium
      • 20-24 (rare strong): apply 50-100%+ premium
    • Per-resolution-economics premium (if cost-per-resolved-task < competitive anchor with sustained GM >65%): +10-25%
    • Intervention-rate trajectory (if declining over 4+ quarters): +5-15%; (if rising): -10-25%
    • Irreversibility governance: clean log + drilled kill-switch + audit-log accepted = +5-10%; recent sev-1 = -15-40%
    • Regulator engagement evidence: pre-clearance / sectoral approval = +5-15%; pending regulatory action = -20-60%
    • Foundation-model platform risk: high (provider actively shipping in your category) = -15-40%; low (your moat survives provider commoditisation) = neutral
    • Multi-agent governance discount: if multi-agent product without branch / loop caps + kill-switch + governance evidence = -20-40%
  4. Compute comparable transactions — recent agent rounds and exits, multiple ranges, adjusted for stage / geography / vertical. Note: agent-comparable data is sparse and noisy in 2026; use carefully.
  5. Apply geographic / market adjustments — African / emerging-market discount per meta-valuation (typically -10 to -30% on SaaS / AI; can be narrower for sovereign-AI-positioned plays).
  6. Output the adjusted multiple range — low / base / high with reasoning per adjustment.
  7. Map to pre-money — adjusted multiple x current ARR (or forward ARR with discount).
  8. Stress the valuation — under provider 5x, irreversibility incident, regulator action, foundation-model commoditisation, AI Safety Lead departure: what does the multiple compress to?
  9. State the investor-archetype-specific multiple — agent-specialist fund / vertical AI fund / generalist SaaS fund / sovereign-AI / DFI all anchor differently.
  10. Wire to the funding ask (11-funding-request) and exit thesis (saas-valuation-and-fundraising-strategy).

Decision, stop, and recovery controls

  • Decision point: confirm that the requested output is the agent valuation overlay and that the decision concerns premium, neutral, or wrapper discount.
  • Stop condition: halt the affected conclusion if required evidence is missing (intervention rate, resolved-task margin, and moat evidence) or if the work could lead to this identified risk: capitalising automation claims that customers or margins do not support.
  • Recovery: obtain the missing record or reviewer, repeat the affected check, and update the exception record before release.

Quality Bar

  • All three layers (SaaS / AI / Agent) adjustments visible and computed transparently
  • Moat-vs-wrapper score directly drives the adjustment, not a vibe
  • Comparable transactions cited with date and source
  • Foundation-model platform risk explicitly priced
  • Geographic adjustment applied
  • Range stated (low / base / high) with reasoning per adjustment
  • Stress on valuation under tail risks
  • Investor-archetype-specific multiples differentiated

Anti-Patterns

  • Quoting headline ARR multiple without adjustments
  • Ignoring wrapper discount because "we use AI"
  • Cherry-picking comparable transactions
  • Ignoring foundation-model platform risk
  • Single-point multiple with no range
  • Mixing SaaS / AI / Agent adjustments without showing each
  • Ignoring geographic discount in emerging markets
  • Pricing on TAM rather than on operating evidence
  • Applying the wrong neighbouring route to meta agent valuation adjustments. Correction: confirm the decision and route to the named neighbour before analysis.
  • Treating an assumption as verified evidence. Correction: label it, cite its source or owner, and assign a verification action.
  • Recommending action without a decision threshold. Correction: state the measurable acceptance condition and review trigger.
  • Recording an unavailable check as passed. Correction: mark it not assessed and state the consequence for the decision.
  • Mutating or publishing during an analysis-only task. Correction: remain read-only until the owner gives explicit authority.

Outputs

ArtefactConsumerObservable acceptance condition
Valuation Adjustments deliverableNamed decision-maker or plan authorThe recommended choice, assumptions, countercase, and next action are explicit.
Evidence and exception registerReviewer, funder, board, or implementation ownerEvery load-bearing claim is sourced or labelled as an assumption; missing checks are not shown as passes.
  • Adjusted multiple range (low / base / high)
  • Adjustment trail (SaaS base -> AI adjustment -> Agent adjustment with each component)
  • Comparable transactions table
  • Pre-money valuation range
  • Stress on valuation under tail risks
  • Investor-archetype-specific multiples
  • Sensitivity to key drivers

Living-Plan Cadence Defaults

ElementCadenceOwnerVariance threshold
Comparable-transactions scanquarterlyHead of Strategynew comparable that moves range
Moat-vs-wrapper rescorequarterly(per saas-agent-moat-and-wrapper-risk)-3 points
Foundation-model platform riskquarterlyCTO + CEOprovider ships in category
Multiple range refreshquarterlyCFO + CEOmultiple drifts >20%
Pre-round valuation setper-roundCEO + Boardround opens

References

  • references/saas-agent-valuation-adjustments.md — adjustment table by archetype and dimension
  • skills/meta-finance/meta-valuation/SKILL.md — SaaS valuation parent
  • skills/meta-finance/meta-ai-valuation-adjustments/SKILL.md — AI valuation parent
  • skills/saas/saas-valuation-and-fundraising-strategy/SKILL.md — SaaS fundraising
  • skills/pipeline/06-competitive-analysis/saas-agent-moat-and-wrapper-risk/SKILL.md — moat score input
  • skills/pipeline/10-financial-projections/saas-agent-unit-economics-and-cogs/SKILL.md — UE input
  • skills/pipeline/12-risk-analysis/saas-agent-risk-and-stress-test/SKILL.md — stress input
  • skills/pipeline/11-funding-request/saas-agent-funding-stage-playbook/SKILL.md — funding consumer
  • skills/pipeline/14-ai-integration/references/ai-agent-sla-layer-archetypes-and-handoffs.md — agent audit

Africa / Uganda Application Notes

  • African discount on agent valuations is typically -15 to -35% on US / EU baseline; partly market-discovery, partly liquidity, partly DFI-co-investment-norms
  • Sovereign-AI positioning premium — agents with credible in-country compute / data residency / regulator engagement in KE / NG / ZA / RW / EG can earn 10-25% premium versus offshore competitors for the same vertical (public-sector procurement preference)
  • African DFI / multilateral co-investment can lower required institutional multiple (DFI patient capital tolerates longer payback); document DFI co-investment intent in the valuation thesis
  • Agent comparable-transactions in Africa are sparse in 2025-2026; rely on global comparables adjusted; document the adjustment explicitly
  • Currency-of-record matters — institutional rounds price in USD; report multiples in USD; report local-currency-equivalent for DFI / strategic
  • Sovereign-AI envelopes (RW innovation envelope; KE Talanta; NG NITDA implementation; ZA Presidential 4IR; EG infrastructure) can be valuation-supportive if executed; document trajectory
  • Liquidity discount — African secondary markets thin; expect 10-15% liquidity discount in pre-money compared to US comparables; reduces at growth-round and trade-sale-eligible stages

Evidence Produced

EvidenceFormatAcceptance condition
Agent valuation overlay decision traceSources, calculations, assumptions, countercase, and selected actionA reviewer can trace the selected action and rejected alternatives to the cited inputs.
Exception recordFailed and not-assessed checks with owner and due actionThe register exposes every unresolved exception that could lead to capitalising automation claims that customers or margins do not support.

Capability and Permission Boundaries

Read supplied records and use non-mutating checks to produce the agent valuation overlay; adding a documented overlay to the valuation workbook is permitted when requested. Do not publish, contact third parties, alter live systems, commit funds, or claim legal, tax, audit, valuation, ESG, or investment assurance without the owner's explicit authorisation and the appropriate reviewer.

Degraded Mode

If intervention rate, resolved-task margin, and moat evidence cannot be obtained, return a qualified agent valuation overlay covering only the checks that remain supportable. Leave this decision unresolved: premium, neutral, or wrapper discount. Record the evidence owner and next check; an inaccessible source, tool, or reviewer is never a pass.

Decision Rules

Decision conditionActionFailure or risk avoided
Evidence is sufficient to decide: premium, neutral, or wrapper discountRecord the conclusion, source trail, owner, and review trigger in the agent valuation overlay.Risk of capitalising automation claims that customers or margins do not support
Material evidence conflicts or remains uncertainShow the valuation with no agent premium and with the evidenced adjustment; use the unsupported case only as sensitivity.Selecting an option without resolving the decision-relevant uncertainty
Required evidence is missing: intervention rate, resolved-task margin, and moat evidenceMark the decision on premium, neutral, or wrapper discount not assessed in the agent valuation overlay, and send it to the valuation lead and transaction adviser.Otherwise, the work risks capitalising automation claims that customers or margins do not support

Quality Standards

Accept the agent valuation overlay only when evidence is sufficient for this decision: premium, neutral, or wrapper discount. Assumptions and countercases remain visible, calculations and cross-references reconcile, and the reviewer can see how the recommendation addresses the risk of capitalising automation claims that customers or margins do not support.

Worked Example

Two agent products use the same foundation model, but only one has proprietary workflow data and falling intervention rates. Value the second case with evidenced adjustments and leave the first at the wrapper case.

Finance Doctrine Gate

Apply the Chwezi doctrine to the agent valuation overlay, using the reporting basis and effective date supported by intervention rate, resolved-task margin, and moat evidence. Reconcile the treatment to the model and narrative, and have the valuation lead and transaction adviser review the treatment, reconciliation, and exposure to this risk: capitalising automation claims that customers or margins do not support.

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

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.