"Learning to Create an Expert Python Program to Exit GitHub" matching MCP tools:

• create_projectA

  KeelenA

  Start building something new: creates a GitHub repo and begins work on it. For a NEW repo. Given an existing repo, import_project(repo_full_name) is the right call — this one would create a second, empty repo beside it (list_github_repos() browses what the workspace can see). Scaffolds a new GitHub repo, a bootstrap-mode project, and submits `build_description` as the project's first Roadmap Request. `name` is a concise GitHub short repo slug (no owner); `project_kind` is REQUIRED and one of library | node_library | python_library | service | cli | web_app | godot_game | roblox_game; `preview_command` is required iff `project_kind == 'web_app'`. `engine` is OPTIONAL — one of claude_code | codex | glm | kimi | grok (defaults to claude_code); codex, glm, kimi, and grok require the workspace to have a matching connected credential. `org` is OPTIONAL — a GitHub organization login to create the repo inside (e.g. your company org); omit it to land the repo on a member's personal account. `private` defaults to True. `ci_runs_on` is OPTIONAL — the CI runner labels for the scaffolded workflow, e.g. ["self-hosted", "linux", "x64", "my-fleet"]. Omit it to inherit the workspace default (ubuntu-latest if unset). Labels no registered org runner carries are rejected, because GitHub would queue such a job forever rather than fail it. `framework` is OPTIONAL and `web_app`-only — one of vite | next (defaults to vite). It picks the scaffolded frontend rails: `vite` a vanilla-TypeScript SPA, `next` a Next.js app-router app. Passing it with any other `project_kind` is an error. The repo is created on the GitHub account of a workspace member with repo-create OAuth access (this path has no specific caller user), so the returned `repo` owner is whichever member's token resolved (or the chosen `org`). If no member has repo-create access — or the resolving member can't create in `org` — the call returns an actionable error. Returns {project_id, repo, thread_id, next_action, poll_after_seconds, next_step}; next_step names the follow-up poll of get_request_status with the returned thread_id. On the rare arm where the first Request failed to submit, next_action is "call_tool" with next_tool="submit_request". If scaffolding fails after the repository has been created, best-effort compensation deletes that just-created repository so a retry can reuse the name — this tool can therefore remove external state it created moments earlier.

  Connector

  Destructive

  No auth

• findagent_list_reposA

  FindAgentA

  List the CALLER'S OWN GitHub repos so you can pick one to import. Returns each repo's full_name (owner/repo), default_branch, and private flag. Pass a chosen full_name to findagent_import_repo. If GitHub is not connected it returns connected:false pointing at findagent_connect_github; in GitHub-App mode with no repos granted it returns needs_install with an install URL. Never returns your GitHub token.

  ConnectorOAuth

• submit_jobA

  MDEngineA

  One call: create a hosted GPU job, upload the deck given INLINE as {relative_path: text}, and queue it. Total inline size <= 8 MB; for larger decks use create_job, PUT the tarball to upload_url, then start_job. Billing starts at the first heartbeat (state running) and stops at done/failed/cancelled. The deck is executed as a program on an isolated GPU pod.

  ConnectorNo auth

• calculate_irr_sensitivityA

  valuation-apiA

  Compute an IRR sensitivity grid across a range of exit multiples and hold periods for a single lump-sum investment. WHEN TO USE: to stress-test how the annualised return varies with exit multiple and holding period before committing to an investment. Complements calculate_irr. WHEN NOT TO USE: when you need one precise IRR for a known exit value (use calculate_irr), or a full valuation (use calculate_dcf). BEHAVIOUR: pure deterministic calculation — no side effects, no network or storage access; idempotent and non-destructive. NOTE ON GRID GEOMETRY: the byMultiple grid is computed at the SECOND hold period in hold_periods (default 5 years); the byHoldPeriod grid is computed at a 2.5x exit multiple. RETURNS: JSON object { byMultiple: { "2.0x": 14.9, ... } with IRR values as percentage numbers rounded to 1dp, byHoldPeriod: { "5y": 18.4, ... } }. PARAMETERS: initial_investment (number > 0), exit_multiples (optional array of numbers to test, default [1.5, 2.0, 2.5, 3.0, 3.5]), hold_periods (optional array of positive integers (years) to test, default [3, 5, 7, 10]).

  ConnectorNo auth

• calculate_irr_sensitivityA

  valuation-apiA

  Compute an IRR sensitivity grid across a range of exit multiples and hold periods for a single lump-sum investment. WHEN TO USE: to stress-test how the annualised return varies with exit multiple and holding period before committing to an investment. Complements calculate_irr. WHEN NOT TO USE: when you need one precise IRR for a known exit value (use calculate_irr), or a full valuation (use calculate_dcf). BEHAVIOUR: pure deterministic calculation — no side effects, no network or storage access; idempotent and non-destructive. NOTE ON GRID GEOMETRY: the byMultiple grid is computed at the SECOND hold period in hold_periods (default 5 years); the byHoldPeriod grid is computed at a 2.5x exit multiple. RETURNS: JSON object { byMultiple: { "2.0x": 14.9, ... } with IRR values as percentage numbers rounded to 1dp, byHoldPeriod: { "5y": 18.4, ... } }. PARAMETERS: initial_investment (number > 0), exit_multiples (optional array of numbers to test, default [1.5, 2.0, 2.5, 3.0, 3.5]), hold_periods (optional array of positive integers (years) to test, default [3, 5, 7, 10]).

  ConnectorNo auth

• calculate_irr_sensitivityA

  Valuation APIA

  Compute an IRR sensitivity grid across a range of exit multiples and hold periods for a single lump-sum investment. WHEN TO USE: to stress-test how the annualised return varies with exit multiple and holding period before committing to an investment. Complements calculate_irr. WHEN NOT TO USE: when you need one precise IRR for a known exit value (use calculate_irr), or a full valuation (use calculate_dcf). BEHAVIOUR: pure deterministic calculation — no side effects, no network or storage access; idempotent and non-destructive. NOTE ON GRID GEOMETRY: the byMultiple grid is computed at the SECOND hold period in hold_periods (default 5 years); the byHoldPeriod grid is computed at a 2.5x exit multiple. RETURNS: JSON object { byMultiple: { "2.0x": 14.9, ... } with IRR values as percentage numbers rounded to 1dp, byHoldPeriod: { "5y": 18.4, ... } }. PARAMETERS: initial_investment (number > 0), exit_multiples (optional array of numbers to test, default [1.5, 2.0, 2.5, 3.0, 3.5]), hold_periods (optional array of positive integers (years) to test, default [3, 5, 7, 10]).

  ConnectorNo auth

• run_codeA

  EQVPS — VPS and code sandboxes for AI agentsA

  Run a code snippet inside a sandbox and return {exit_code, stdout, stderr, timed_out, truncated, duration_ms}. `language`: python (Python 3.12), node (Node.js 22) or bash. Install packages first with exec_command (`pip install …` / `npm i …`). The code is passed to the interpreter on stdin, so it cannot read stdin itself. `timeout` seconds (default 30, max 55 per call). LONGER JOBS: set `background: true` — returns a `task_id` immediately (timeout then defaults to the sandbox's remaining lifetime); poll get_task with the returned offsets to read new output and the exit code, kill_task to stop it. The sandbox keeps billing while a task runs and is not idle-deleted. Output is capped at 1 MiB per stream for synchronous calls. Files and installed state persist between calls in the same sandbox. Requires: token + live sandbox_id. Typical: timed_out after 55 s → repeat with background: true.

  ConnectorNo auth

• handoffs.agencyA

  AI Design BlueprintA

  Authenticated: submit an agency engagement enquiry on behalf of the caller for a founder-led discovery call. Persists an AgencyHandoff row routed to the agency inbox; the user is contacted by the team for a scoped proposal. Engagement scopes: workflow sprint (rapid agentic workflow implementation), proof-of-concept (validate a specific agent design in a bounded timeframe), pilot support (co-design and validate a production-ready pilot), advisory (ongoing architectural guidance across a product team). WHEN TO CALL: the user has identified a paid hands-on expert engagement need beyond self-service learning, and explicitly asks to talk to the team or book a discovery call. ALWAYS confirm with the user before firing: this creates a sales-visible record. WHEN NOT TO CALL: for free training / partnerships discussion (use handoffs.partnership); for support / billing / access (use handoffs.operator); proactively or as a sales push. BEHAVIOR: write-only, single insert, side-effecting. Auth: Bearer <token> (any plan). Data at rest in the UK. Response confirms the ticket id + scope so the user can reference it.

  ConnectorNo auth

• contacter_expertA

  payotte-mcpB

  Call this ONLY when the user explicitly asks to contact, reach out to, or request a quote/appointment from a Payotte-listed professional. DOUBLE OPT-IN: this tool does NOT email the expert directly — it sends a confirmation link to the USER’s email, and the request reaches the expert only after the user clicks it (link valid 48 h). Tell the user to check their inbox. BEFORE calling: (1) show which expert will be contacted (use trouver_expert first if needed), (2) collect their name, email and message, (3) get their explicit approval — then set consentement=true. Never invent contact details. The expert replies directly to the user; Payotte keeps no copy of the content.

  ConnectorNo auth

• list_user_lp_positionsA

  SwapWizardA

  Maps to GET /positions. Reads all LP positions a wallet holds on a given chain by calling the SwapWizard API, which discovers positions across all supported protocols: Uniswap V2/V3/V4, Aerodrome, Thena, SushiSwap, PancakeSwap, Algebra, Balancer, Curve, and all Solidly forks. Each position includes positionId, nftManager, dexName, liquidityKind, token addresses, amounts, fees, in-range status, APR, and USD values. EXIT SIGNAL: each position also carries `momentumSignal` for its pool — "exit" means the pool's volume is dying or LPs are leaving (consider zapping out), "watch"/"entry" mean momentum is still alive. When `momentumSignal` is ABSENT, the pool has dropped out of the momentum ranking (momentum exhausted) — also a reason to review and likely exit the position. `momentumScore` is the composite strength. Use this to drive exit decisions, mirroring the entry signal from search_liquidity_pools. The API uses Alchemy's NFT APIs for optimal position discovery — pass an Alchemy RPC URL via rpcUrl for fastest results. Without an Alchemy key, the API falls back to on-chain scanning which may be slower and newly created positions may take longer to appear. IMPORTANT: Always call this BEFORE zap_out_of_lp_position — pass the returned positionId, nftManager, dexName, and liquidityKind directly to zap_out_of_lp_position.

  ConnectorNo auth

• handoffs.agencyA

  AI Design Blueprint DoctrineA

  Authenticated: submit an agency engagement enquiry on behalf of the caller for a founder-led discovery call. Persists an AgencyHandoff row routed to the agency inbox; the user is contacted by the team for a scoped proposal. Engagement scopes: workflow sprint (rapid agentic workflow implementation), proof-of-concept (validate a specific agent design in a bounded timeframe), pilot support (co-design and validate a production-ready pilot), advisory (ongoing architectural guidance across a product team). WHEN TO CALL: the user has identified a paid hands-on expert engagement need beyond self-service learning, and explicitly asks to talk to the team or book a discovery call. ALWAYS confirm with the user before firing: this creates a sales-visible record. WHEN NOT TO CALL: for free training / partnerships discussion (use handoffs.partnership); for support / billing / access (use handoffs.operator); proactively or as a sales push. BEHAVIOR: write-only, single insert, side-effecting. Auth: Bearer <token> (any plan). Data at rest in the UK. Response confirms the ticket id + scope so the user can reference it.

  ConnectorNo auth

• list_user_lp_positionsA

  SwapWizard MCPA

  Maps to GET /positions. Reads all LP positions a wallet holds on a given chain by calling the SwapWizard API, which discovers positions across all supported protocols: Uniswap V2/V3/V4, Aerodrome, Thena, SushiSwap, PancakeSwap, Algebra, Balancer, Curve, and all Solidly forks. Each position includes positionId, nftManager, dexName, liquidityKind, token addresses, amounts, fees, in-range status, APR, and USD values. EXIT SIGNAL: each position also carries `momentumSignal` for its pool — "exit" means the pool's volume is dying or LPs are leaving (consider zapping out), "watch"/"entry" mean momentum is still alive. When `momentumSignal` is ABSENT, the pool has dropped out of the momentum ranking (momentum exhausted) — also a reason to review and likely exit the position. `momentumScore` is the composite strength. Use this to drive exit decisions, mirroring the entry signal from search_liquidity_pools. The API uses Alchemy's NFT APIs for optimal position discovery — pass an Alchemy RPC URL via rpcUrl for fastest results. Without an Alchemy key, the API falls back to on-chain scanning which may be slower and newly created positions may take longer to appear. IMPORTANT: Always call this BEFORE zap_out_of_lp_position — pass the returned positionId, nftManager, dexName, and liquidityKind directly to zap_out_of_lp_position.

  ConnectorNo auth

• request_callbackA

  PropContext — Bangalore Real EstateA

  Request an expert callback for a Bangalore real estate project. Use this when the user wants to speak to an expert but is not ready to visit yet. Lower commitment than a site visit — just captures name, phone, and preferred call time. Call this when the user says: - "I want to know more", "can someone call me", "I'd like a callback" - "talk to an expert", "get more information", "not ready to visit yet" - "call me back", "have someone reach out" Required: rera_number, user_name, user_phone Optional: preferred_time (e.g. 'Morning', 'Afternoon', 'Evening') conversation_summary: Summarise in 2-3 bullet points what the buyer is looking for and any questions they raised — this goes to the expert who calls them back. Returns confirmation with callback reference ID.

  ConnectorNo auth

• findagent_create_code_draftA

  FindAgentA

  Persist a CODE-BUNDLE draft from YOUR OWN GitHub repo — for an agent that ships RUNNABLE code (use this when findagent_import_repo returned grounding.code_bundle). Pass the basics (title/slug/tagline/description/category_slug + example_prompts: 1-5 required) + the detected contract from import_repo's grounding.code_bundle (runtime, entrypoint {path,export}, mcp {mode,command,args}, ui {path}, allowed_hosts, credential_slots, skills), overriding any you want to correct. The server RE-PULLS the repo (your stored GitHub token — private repos work, server-side), snapshots + scans the code, validates the manifest, and creates a status=draft agent you own; then call findagent_submit_for_review IN THIS MCP CLIENT to set price + confirm originality/prohibited + submit it (the web is only an optional preview). IDEMPOTENT BY REPO: if you already have a draft for this repo, calling this again OVERWRITES that same draft (basics + manifest + a fresh re-pull/re-scan) instead of creating a duplicate — so iterate freely (the response `updated` flag is true on overwrite). NEVER send secret credential VALUES — credential_slots declare shape (ref/env/label/allowed_hosts/type) only. Building/running the code stays gated until an admin approves it. Which tool for which part: Instructions, Skills and Actions are authored as an Agent Plugins package and sent as files with findagent_create_package_draft; Code (a Node or Python program) uses findagent_create_code_draft from your own GitHub repo; an MCP server you already run uses findagent_create_remote_mcp (hosted over https) or findagent_create_local_mcp (started on the buyer machine by a command); an OpenAPI spec becomes tools through findagent_import_openapi before findagent_create_draft; findagent_create_draft saves a declarative listing from findagent_import_repo grounding. Code part: start from https://github.com/FindAgent/agent-template (Node and Python variants). Building any part with an assistant: provider-neutral playbooks at https://github.com/FindAgent/agent-creators (start with AGENTS.md).

  ConnectorOAuth

• findagent_create_package_draftA

  FindAgentA

  Create a DRAFT agent from an Agent Plugins package you wrote in this conversation: pass `files` as an object mapping each path to its TEXT (for example "skills/my-skill/SKILL.md", "skills/my-skill/references/guide.md", ".claude-plugin/plugin.json", "commands/review.md", "agents/reviewer.md"). FindAgent imports it exactly like an uploaded package — the same skill rules (at most 40 skills, valid Agent Skills names; too many is refused, never cut), the same caps (4 MB of text in total, 2 MB per file, 4000 files; the whole request, JSON escaping included, must also stay under the hosting platform's ~4.5 MB request limit), actions only where plugin.json declares them — stores it in your own namespace and saves a status=draft skills agent you own. Binary files cannot be sent. You must pass attestation: "own-work", confirming these are your own files. Then call findagent_submit_for_review in this client (price, originality and prohibited-content confirmations): the package is scanned and a person reviews it before anything is published. Instead of `files` you may pass `repo` (owner/repo or a github.com URL) and optionally `ref` and `path` (the folder of a repository that holds several packages): FindAgent then reads the package from a GitHub repository YOU OWN (it proves you have push access first; connect GitHub with findagent_connect_github if needed), and no attestation is needed because the ownership is proven. Pass exactly one of `files` or `repo`. An agent made from a repository is updated by re-pulling it (findagent_repull). Listing fields are the same as findagent_create_draft (title, slug, tagline, description, category); the instructions, prompts and actions come from the files. NEVER put secret values in any file: an action's credentials are declared by name only. Which tool for which part: Instructions, Skills and Actions are authored as an Agent Plugins package and sent as files with findagent_create_package_draft; Code (a Node or Python program) uses findagent_create_code_draft from your own GitHub repo; an MCP server you already run uses findagent_create_remote_mcp (hosted over https) or findagent_create_local_mcp (started on the buyer machine by a command); an OpenAPI spec becomes tools through findagent_import_openapi before findagent_create_draft; findagent_create_draft saves a declarative listing from findagent_import_repo grounding. Code part: start from https://github.com/FindAgent/agent-template (Node and Python variants). Building any part with an assistant: provider-neutral playbooks at https://github.com/FindAgent/agent-creators (start with AGENTS.md).

  ConnectorOAuth

• findagent_reuploadA

  FindAgentA

  Publish a NEW version of YOUR OWN published skills agent that was UPLOADED (not imported from GitHub), from files you send as text: `files` maps each path to its text, exactly as findagent_create_package_draft takes them, with the same skill rules and caps. You must pass attestation: "own-work". FindAgent imports the files, diffs them against your live version and, if they changed, submits a new version for review with an auto-drafted changelog; your LIVE version keeps serving until it is approved. Unchanged files are a no-op. An agent imported from GitHub is refused (`github_source`): it re-versions by re-pulling its repository with findagent_repull, because an attested upload must never replace proven provenance. Optional bump (patch|minor|major, default patch). Which tool for which part: Instructions, Skills and Actions are authored as an Agent Plugins package and sent as files with findagent_create_package_draft; Code (a Node or Python program) uses findagent_create_code_draft from your own GitHub repo; an MCP server you already run uses findagent_create_remote_mcp (hosted over https) or findagent_create_local_mcp (started on the buyer machine by a command); an OpenAPI spec becomes tools through findagent_import_openapi before findagent_create_draft; findagent_create_draft saves a declarative listing from findagent_import_repo grounding.

  ConnectorOAuth

• save_program_to_phormeA

  Phorme TrainingA

  Use this when the user has seen a program preview and clearly said to save it. It saves that preview to the connected Phorme account as a private program and returns a link that opens it in Phorme. It needs a connected Phorme account with the training:write permission and a Phorme membership that already includes programs. If the account does not have that membership the tool says so and saves nothing. It never starts a purchase. Retrying with the same idempotency_key does not create a second copy.

  ConnectorNo auth

• program_outcome_priorA

  vuln-intelA

  At the START of an engagement on a bug-bounty PROGRAM, get a grounded prior over what bug classes have historically LANDED there — to decide what to look at first and what to skip. Built from real HackerOne disclosed reports: the class distribution for this program, each with LIFT over the global base-rate, sample size, and cited example reports; plus `lands_rarely` — the classes that land broadly across programs but almost never here (where NOT to spend the first hour). DESCRIPTIVE, not predictive: it reports what *landed*, never claims what you *will* find. It's a disclosure-selected sample, HackerOne-only (~379 programs with public disclosures), models bug-class not payout, and withholds a prior when the program has too few reports (an anecdote isn't a distribution). Pass a program handle or name, e.g. "gitlab", "shopify". Complements find_attack_approaches (mechanism transfer) and hunt_plan (CVE dig-order) — this is the minute-zero "where to start" prior an agent has no tacit target knowledge to form on its own.

  ConnectorNo auth

• send_feedbackA

  OpenCancerA

  Flags an OpenCancer answer for human expert review, or sends feedback to the OpenCancer team. Used when the user says an OpenCancer answer was wrong, unsafe, or incomplete, or wants to tell OpenCancer something (answer_id identifies the answer; message holds the user's words). Also used to report that an OpenCancer tool returned no usable answer (kind "missing_answer" with answer_id or tool, no message). Feedback is read by the OpenCancer team; it is not an urgent-care channel and no reply is sent through it.

  ConnectorNo auth

• list_expert_profilesA

  TasteA

  Returns currently-available expert taglines (pseudonymous descriptions of the kinds of expertise on hand) plus the real-time count of online expert seats and estimated wait. Use this as a cheap pre-flight check before calling a paid tool. Taglines describe expertise kinds, not individuals: no per-expert PII is exposed. Free.

  ConnectorNo auth

• list_domainsA

  mcp-serverA

  List the taxonomy domains the company has indexed — with document counts, expert counts, and coverage levels — so an agent can decide whether to query before spending a Knowledge Token. Returns one row per domain with the canonical `taxonomy_domain` slug, document/chunk counts, expert count, coverage level (expert | partial | none), the single_expert risk flag, and the top contributor by authority. Use the slug as the `domain` filter on a follow-up `query_knowledge` call. Zero Knowledge Tokens consumed.

  ConnectorNo auth