priiism_container_status | Get the current container/sandbox status including dev server readiness, uptime, and tunnel URL. | None |
priiism_preview_url | Get the live preview URL for this project. Returns the tunnel URL that serves the running dev server. | None |
priiism_project_info | Get project metadata including name, description, status, template, and timestamps. | None |
priiism_config_get | Read the viiibin.config.json for this project. This config file is REQUIRED for deployments. Returns the parsed config or an error if missing. The config schema is: { version: 1, name: string, description?: string, targets: Array, configVariables?: Record<string, ConfigVariable>, configFiles?: Array, postCreate?: { install?, migrate?, seed? } }. Target types: “web” (build-based web deploy, fields: buildCommand?, outputDir?), “static” (no build, fields: buildCommand?, outputDir?), “serverless” (fields: entryPoint?, routes?), “capacitor-ios” (fields: bundleId required, appName?, webDir?, minIosVersion?), “capacitor-android” (fields: bundleId required, appName?, webDir?, minSdkVersion?, targetSdkVersion?). | None |
priiism_config_set | Write/update the viiibin.config.json for this project. This config is REQUIRED for deployments. You must provide the full config object. Minimal example for a static site: { “version”: 1, “name”: “My App”, “targets”: [{ “type”: “static”, “name”: “web” }] }. For a build-based site: { “version”: 1, “name”: “My App”, “targets”: [{ “type”: “web”, “name”: “web”, “buildCommand”: “npm run build”, “outputDir”: “dist” }] }. | config (required) |
priiism_file_read | Read the text content of a single project file by its path (relative to the project root, e.g. “src/index.ts”). Reads from durable project storage (archive-aware), so it works without a running sandbox. Returns the path and content, or an error if the file does not exist. | path (required) |
priiism_file_write | Create or overwrite a single project file with the given text content. Writes to durable project storage and tracks the file in the project. The path is relative to the project root. Overwrites existing files. | path (required), content (required) |
priiism_file_list | List the paths of all files in the project (relative to the project root). Optionally filter to a path prefix. Returns an array of file paths. | prefix |
priiism_file_delete | Delete a single project file by its path (relative to the project root). Returns an error if the file does not exist. | path (required) |
priiism_project_list | List all projects in the authenticated organization (excluding deleted ones). Returns an array of projects with id, name, slug, description, status, templateId, and timestamps. The organization is taken from the API key/token — you do not pass it. | None |
priiism_project_create | Create a new project in the authenticated organization. Provide a name; optionally a description, a templateId to scaffold from a template (omit or use ‘empty’ for a blank project), and configValues supplying that template’s configuration variables. Returns the created project (and fileCount when a template populated files). | name (required), description, templateId, configValues |
priiism_project_update | Update an existing project’s name and/or description. Identify the target project by projectId. At least one of name or description must be provided. Returns the updated project. | projectId (required), name, description |
priiism_template_list | List the active project templates available to the authenticated organization. Returns an array of templates with id, name, description, category, and isFeatured. | None |
priiism_template_instantiate | Create a new project from a template. Provide the templateId to instantiate and a projectName for the new project; optionally configValues supplying the template’s configuration variables. Returns the created project (and fileCount when the template populated files). | templateId (required), projectName (required), configValues |
priiism_actions_list | List the pending AI actions awaiting approval for this project. Optionally filter by status (defaults to “pending”). Returns an array of pending actions with id, toolName, description, impact, category, status, and batch grouping. | status (pending |
priiism_action_approve | Approve a single pending action by its id. Optionally set execute to run it immediately after approval. Returns the action id and, when executed, the execution result. Fails if the action is not currently pending. | actionId (required), execute |
priiism_action_reject | Reject a single pending action by its id, with an optional reason. Fails if the action is not currently pending. | actionId (required), reason |
priiism_action_batch | Approve or reject every still-pending action in a batch by its batchId. Returns the batchId, the decision, and the number of actions resolved (0 on a retry after the batch is already resolved). | batchId (required), decision (required; approve |
priiism_job_estimate | Estimate the credit + dollar cost of delivering one or more milestones. Provide either a milestoneUrl, or a repo (“owner/name”) plus a milestones array of milestone numbers (or selectAll to estimate every open milestone). Returns a per-milestone cost breakdown and an aggregate total. Read-only — computes an estimate without queuing anything. | milestoneUrl, repo, milestones, selectAll |
priiism_job_execute | Release a queued job for pickup by the background delivery runner. Returns 202 with the job id and queued status — it does NOT run the job synchronously; poll job_status for progress. Fails if the platform is not configured to run jobs. | jobId (required) |
priiism_job_status | Get one job plus its per-issue work items (ordered by execution order). Returns the job record and its items array. | jobId (required) |
priiism_job_pause | Pause a running job. Returns the job id and its new paused status. Fails if the job is not currently running. | jobId (required) |
priiism_job_resume | Resume a paused job, optionally raising its cost or token budget. Returns the job id and its new running status. Fails if the job is not currently paused. | jobId (required), newCostLimit, newTokensLimit |
priiism_plan_propose | Kick off a planner run against the project’s latest succeeded knowledge build. Returns the newly-created pending/running proposal, or a 409 with the existing in-flight proposal if one is already running. Optionally pass a model to override the default. | model |
priiism_plan_dryrun | Run a reconcile preview for a proposal against the linked repository without writing to it. Returns the decision list (which milestones/issues would be created, attached, suffixed, or skipped) plus the judgement cost. Read-only. | proposalId (required) |
priiism_plan_push | Apply a proposal to the linked repository: create or attach milestones and issues. Supports perMilestoneOverrides to force a per-milestone policy (create/attach/suffix/skip). Partial pushes leave the proposal un-pushed so a retry re-attempts only the missing items. | proposalId (required), perMilestoneOverrides |
priiism_proposals_list | List the plan-proposal history for this project (paginated). Optionally pass limit and an opaque cursor. Returns the proposals array and a nextCursor. | limit, cursor |
priiism_security_scan | Run a security scan over the project’s current files: static analysis (SAST) plus a dependency-vulnerability audit. Persists a scan record and returns { scanId, status, score, findingsCount, criticalCount, highCount }. Each call runs a fresh scan. | None |
priiism_security_findings | List security findings from the most recent scans, newest first. Optionally filter by status (“open" | "resolved" |
priiism_domain_bind | Bind a verified custom domain to the project’s live web hosting. The domain must already be verified (call domain_verify first) — binding an unverified domain returns a 409. Returns the updated domain record. | domainId (required) |
priiism_domain_verify | Check whether a custom domain’s ownership DNS record is in place and, if so, transition it to verified. Returns { verified, domain }. | domainId (required) |
priiism_domain_status | Poll the live status of every custom domain on the project (refreshing each from the hosting platform) and return the current domain list. Read-only. | None |
priiism_identity_get | Read the project’s managed identity-provider configuration: provider, provisioning status, callback URLs, allowed logout URLs, and web origins. Never returns the provider client secret. Read-only. | None |
priiism_identity_configure | Provision or update the project’s managed identity-provider configuration (callback URLs, web origins, optional logout URLs). Upserts — safe to call repeatedly. Returns the same shape as identity_get. | provider (required), callbackUrls (required), webOrigins (required), allowedLogoutUrls |
priiism_conversations_list | List the project’s chat conversations (id, title, createdAt), newest first. Optionally pass limit (default 20, max 100). Read-only. | limit |
priiism_org_members_list | List the members of the API key’s organization (userId, email, name, role). Read-only. | None |
priiism_autonomy_start | DANGEROUS — requires the opt-in autonomy:run scope, which is OFF by default and is NEVER included in any role bundle; a key must carry it explicitly. Starts the autonomous multi-step agent loop for the project’s latest succeeded plan proposal (or an explicit proposalId). This SPENDS credits and opens real pull requests. If a run is already in progress for the resolved proposal, returns a 409 with the existing runId instead of starting a duplicate. | proposalId, stepMode |
priiism_autonomy_status | Read the status of an autonomous run: returns the plan_autonomy_runs row plus the most recent plan_node_runs for the run’s proposal. Read-only. Requires the autonomy:read scope. | runId (required) |
priiism_seos_roles_list | List the project’s SEOS engineering-role charters (worker/reviewer/release/lead tiers and their owned levers). Read-only. Requires the autonomy:read scope. | None |
priiism_container_exec | DANGEROUS — requires the opt-in container:exec scope, which is OFF by default and is NEVER included in any role bundle; a key must carry it explicitly. Runs an arbitrary shell command in the project’s active sandbox. The command is validated against the platform command allowlist BEFORE execution — blocked commands (e.g. rm -rf /) return an error and never run. Returns { stdout, stderr, exitCode } with each stream truncated at 10,000 characters. | command (required), timeout, workdir |
priiism_code_search | DANGEROUS — requires the opt-in code:search scope, which is OFF by default and is NEVER included in any role bundle; a key must carry it explicitly. Runs a RAG semantic search over the whole project’s indexed codebase and returns the most relevant chunks (filePath, content, score, startLine, endLine). | query (required), topK, minScore |