Borrowing it
Nothing to install: this file belongs to ElioLin/sfoa-enterprise-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ElioLin/sfoa-enterprise-mcp/main/.codebuddy/skills/sfoa-salesforce-assistant/SKILL.mdgit clone --depth 1 https://github.com/ElioLin/sfoa-enterprise-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/eliolin/sfoa-enterprise-mcp/sfoa-salesforce-assistant)<a href="https://agentmods.dev/skills/eliolin/sfoa-enterprise-mcp/sfoa-salesforce-assistant"><img src="https://agentmods.dev/badge/skills/eliolin/sfoa-enterprise-mcp/sfoa-salesforce-assistant/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/eliolin/sfoa-enterprise-mcp/sfoa-salesforce-assistant"><img src="https://agentmods.dev/badge/skills/eliolin/sfoa-enterprise-mcp/sfoa-salesforce-assistant.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00039 | $0.01887 |
| Opus 5 | $0.00019 | $0.00944 |
| Sonnet 5 | $0.00008 | $0.00377 |
| Haiku 4.5 | $0.00004 | $0.00189 |
Grade A, and why
sfoa-salesforce-assistant 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 today.
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.
How it starts
The opening of the file, as written. The whole thing — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SFoA Salesforce Assistant
Canonical Playbook version: 1.8.0.
When to use
Use this Skill when a user asks for current Salesforce business data, an allowed CREATE/UPDATE, Salesforce behavior diagnosis, Lookup/Picklist resolution, or a usable record link.
Required workflow
- Read references/tool-workflows.md and select only a workflow supported by the Connector's current MCP capabilities.
- Before mutation or diagnosis, read references/safety-boundaries.md.
- Obtain current capability facts from
sfoa://agent-capabilities/currentwhen the Connector supports Resources. - If Resources are unavailable and
get_agent_playbookis exposed, use that Tool fallback. Never call an absent Tool.
SFoA Salesforce Agent Playbook
Playbook-Version: 1.8.0 Workflow: ALL
Runtime capabilities
- This is a distribution template. Discover current Tools and policy from MCP; no capability is implied by this file.
- Dynamic Forms evidence: runtime/object-policy dependent; treat it as
NOT_AVAILABLEunless the current CREATE context includes effective field facts.
BATCH — Bounded synchronous batch mutation
- Tool selection matrix — choose by request size and by the Tools actually advertised, never by the Tool you assume exists. Both mutation Tools enabled: exactly 1 record prefers
create_record/update_record, and 2..200 independent records of one object prefercreate_records/update_records. Plural Tool disabled: 1 record uses the singular Tool and 2+ records use bounded singular calls. Singular Tool disabled: 1 record uses the plural Tool carrying exactly 1 item, and 2..200 records use it normally. Both disabled: that operation is unavailable. Never call a Tool that the current connection does not advertise as enabled. Batch Tools retain every single-record allowlist, managed-field and request USER rule. Never use DELETE or automatic UPSERT. - A clear request to create, modify, or change all matching records already supplies mutation intent. Do not add confirmation merely because a batch is involved. Ask only for ambiguous targets/Record Types/Lookups, missing required values, incomplete scope, or unclear execution intent.
- Reuse action context for the same object, operation, Record Type and relevant field semantics. Prefer grouping by object + Record Type + relevant context. Different draft values that change Dynamic Forms, required dependencies or dependent Picklists require their own refinement; do not make identical metadata calls for every row.
- Each CREATE item carries its resolved recordTypeId and latest uiContextResolutionId when supplied. Never mix the Record Type of collected context with a different mutation RecordTypeId. DML fields always use API values and proven Lookup IDs.
- Use allOrNone=false by default for independent business records; allOrNone=true may protect a strongly related collection within one request. Neither setting creates a transaction across Tool calls or batches.
- For all matching records, prove the complete target scope before mutation: LIMIT, truncation, timeout, pagination or a missing completion indicator cannot prove all records were found. Continue bounded reads/pagination to establish completeness or stop and ask about scope. Never silently mutate a subset and report all complete.
- For an explicit scope over 200 records, split into bounded calls of at most 200, for example 500 becomes 200 + 200 + 100. Fix a finite plan from the known total; track total/processed/succeeded/failed/unknown throughout. Each batch is independent with no global allOrNone. Stop automatic continuation on any OUTCOME_UNKNOWN and verify state first; never loop indefinitely.
- Report SUCCESS only when all requested items have proven success. PARTIAL_SUCCESS means some items succeeded and some failed, and the successful items are already committed. Never resubmit the original batch after PARTIAL_SUCCESS: treat every successful item as already created or updated. Only when the user intent still requires the remaining work and the failure cause is fixable may you prepare a new batch containing exactly the FAILED items. If any result is OUTCOME_UNKNOWN, do not automatically retry anything. Preserve clientReferenceId -> Salesforce recordId mapping from successful results; clientReferenceId is only correlation, never a Salesforce business field or idempotency key.
- If the plural Tools are disabled, bounded singular calls may fulfill the same authorized business intent. If the singular Tools are disabled, a single-record request is satisfied by the plural Tool carrying exactly 1 item. Do not call an absent Tool or claim the entire business request is unsupported merely because one execution shape is unavailable.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- today Changed · +1 lines 4994f58f32a3
- 2d ago Changed · +38 lines b21fda2b12cb
- 4d ago Changed f0608a5f0126
- 8d ago Changed bdf17460de90
- 11d ago First seen · 38 lines · 39 tokens per session scan A 24821707a09c
sfoa-salesforce-assistant is a skill published in the GitHub repository ElioLin/sfoa-enterprise-mcp (0 stars, last pushed today), licensed Apache-2.0. It adds 39 tokens to every session and 1,887 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 2026-08-31.
Other skills, from other repositories
orbit-notion
Open Orbit briefing skill — selected by the Orbit pipeline when Notion is the user's only connected connector, or when the user explicitly scopes their daily digest to Notion. Pulls the past 24 hours of document edits, comments, mentions, and database row changes from the user's authenticated Notion connection and…
agentmail
Use your assigned AgentMail inbox to read email tasks, explicitly send or reply, and check delivery. Provided automatically by your inbox assignment.
Cortex
Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
peekaboo
Capture and automate macOS UI with the Peekaboo CLI.