agentleFS
harshsinghmpharshsinghmp/muse-skills/skills/agency-delivery/ops/SKILL.md

ops

Skill1 starsChanged 17 days agoMITView on GitHub

Internal agency operations department: client onboarding, proposals, statements of work, productized agency SKU packaging and value triangulation, shaping an ambiguous ask and setting its appetite, milestone tracking, project retrospectives, multi-client portfolio management, multi-client context vault and asset fencing, vendor management, Obsidian PKM vault workflows, and agency legal templates (MSAs, subcontractor IP, NDAs, and AI disclosures) — routed through sixteen modes. Use when asked to onboard a new client, write a proposal or SOW, package productized agency service SKUs, triangulate outcome-based pricing, shape an unsized or grab-bag ask before it is priced, track project milestones and scope, run a project retro, manage across multiple client engagements, enforce client context vault isolation, manage vendors, interact with Obsidian knowledge vaults, or draft agency legal agreements. Not for client-facing delivery work (design/webdev/seo) or session handoffs between agents (handoff).

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 14 edited versions we read keep these.

  • 2

    Distinctive

    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:

    Every invocation resolves to exactly one mode.

    line 38
  2. Distinctive:

    Only the resolved mode's reference is loaded — the rest stay on disk, saving tokens on every run.

    line 57

“Common” is measured: kept by at least 6 in 10 of the 14 edited versions of this file we could read. “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

🗂️ ops — Agency Operations Department

One head skill for agency operations. Protect both the client's outcome and the agency's margin. Scope is a promise: write it down, track it, and change it only through an explicit change order. Every client gets a clean workspace, a clear plan, and an honest retrospective.


Modes

Every invocation resolves to exactly one mode. Match the request, then load only the matched reference:

ModeTrigger phrasesBehaviorReference
onboarding"onboard a client", "new client", "kickoff", "client intake"Client intake and isolated workspace setupreferences/onboarding.md
proposal"write a proposal", "pitch a client", "win this project", "proposal doc"Client proposal that connects outcome to plan and pricereferences/proposal.md
sow"statement of work", "sow", "scope document", "contract scope"Statement of work: scope, deliverables, terms, change processreferences/sow.md
sku"productized sku", "sku packager", "turnkey service", "fixed-scope packaging", "value pricing triangulation", "sow generator", "client intake brief", "milestone sign-off"Turnkey productized agency SKU packager, outcome-based value triangulation (5.0x ROI floor), automated SOW and intake brief generation, and milestone verification packetsreferences/sku.md
milestone"track milestones", "scope creep", "project tracking", "delivery status"Milestone and scope tracking with change controlreferences/milestone.md
shape"shape this ask", "set an appetite", "is this sized yet", "grab-bag project", "unsized request"Shape an ambiguous ask before it is priced: named problem, appetite band, no-gos, de-risked pitchreferences/shape.md
retro"retro", "retrospective", "project review", "what went wrong"Blameless project retrospective with tracked actionsreferences/retro.md
multi-client"multiple clients", "portfolio", "manage all clients", "capacity planning"Multi-client portfolio management and capacity planningreferences/multi-client.md
vault"client vault", "context isolation", "asset fencing", "cross-client bleed", "worktree lease", "credential masking"Multi-client context vault, brand asset fencing, shared worktree lease arbitration, and multi-tenant credential maskingreferences/vault.md
contracts"draft a contract", "MSA", "NDA", "legal doc"Legal document and contract drafting standardsreferences/contracts.md
vendor"vendor", "subcontractor", "freelancer management", "procurement", "tooling spend", "manage vendors"Vendor/subcontractor scorecard, spend control, and tooling consolidationreferences/vendor.md
obsidian"obsidian", "obsidian vault", "obsidian cli", "obsidian markdown", "wikilinks", "callouts", "vault notes", "daily note"Obsidian PKM vault operations, wikilinks, callouts, embeds, frontmatter properties, CLI commands, and plugin devreferences/obsidian.md
legal"agency contract", "msa", "subcontractor agreement", "ip assignment", "nda", "legal templates", "privacy policy"Master Services Agreements (MSAs), subcontractor IP assignments, NDAs, and AI compliancereferences/legal.md
audit"audit ops", "process audit", "documentation audit", "milestone audit"Process audit (proposals, SOWs, milestones) + documentation audit (cross-refs resolve)references/audit.md

Only the resolved mode's reference is loaded — the rest stay on disk, saving tokens on every run.


Support references (load on topic)

Loadable by topic; also routable as /ops-<file> commands (not separate routed modes):

  • meeting-notes.md
  • product-marketing-template.md

When to Use

  • Onboarding a new client into a clean workspace.
  • Writing a proposal to win an engagement.
  • Drafting a statement of work with scope and terms.
  • Shaping an ambiguous ask before anyone prices it: appetite band, named problem, no-gos.
  • Tracking milestones, scope, and change orders.
  • Running a project retrospective.
  • Managing a portfolio of concurrent clients.
  • Managing and structuring Obsidian PKM knowledge vaults, wikilinks, callouts, and frontmatter properties.
  • Automating notes, tasks, properties, daily logs, and plugin verification via the Obsidian CLI.

Anti-Triggers

  • Client-facing delivery (design, code, SEO) → the relevant department skill.
  • Session handoff between agents → handoff.
  • Internal quality and evidence tracking → secretary / evidence-ledger.
  • Billing/accounting software setup → out of scope.

Quick Reference

Engagement ladder (decide before any mode)

EngagementDeliverable spine
Discovery/consultFindings doc + recommendation (no build)
Fixed-scope projectSOW + milestone plan + change-order process
RetainerMonthly scope + cadence + reporting
Multi-client portfolioPer-client workspace + cross-portfolio dashboard

Rule: match the paperwork to the risk. Big ambiguous work needs a discovery phase before a fixed price.

Verification gate (every mode)

  • Scope is written, dated, and agreed by both sides.
  • Deliverables have owners and dates.
  • Changes go through a change order, not a verbal 'sure'.
  • Client workspace is isolated and named clearly.

Suite contracts

  • Isolated client workspace scaffold → new-project (client mode).
  • Session continuity across the engagement → handoff / context-anchor.
  • Evidence of work done → evidence-ledger.

Procedure

  1. Intake. establish the client, the engagement type, and what success looks like for both sides — before promising anything; unwritten scope is how projects lose money.
  2. Resolve the mode. Match the request against the mode table; exactly one mode. Ambiguous → ask one question, then proceed.
  3. Execute the mode playbook. Load references/<mode>.md and follow its Intake → Deliverable → Procedure → Quality gate in order.
  4. Gate and deliver. Pass this file's Verification checklist plus the mode's quality gate, then deliver the artifact where the client expects it.

Pitfalls

  • Starting work before scope is written and signed.
  • Verbal scope changes with no change order — margin evaporates.
  • Proposals priced before the work is understood.
  • Client documents mixed into the agency's own workspace.
  • Milestones with no owner or date.
  • Retrospectives that blame people instead of fixing process.
  • Multi-client context bleed (one client's details in another's work).

Verification

  • Scope written, dated, and agreed before work starts.
  • Deliverables have owners and dates.
  • Change-order process defined and used.
  • Client workspace isolated and clearly named.
  • Retrospectives are blameless with tracked actions.
  • No cross-client context leakage.

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.