@pikaa-ai/pikaa 0.2.5 → 0.3.0

package/skills/tamarind/SKILL.md

@@ -0,0 +1,285 @@
+---
+name: tamarind
+description: Access a collection of open-source molecular design and structural biology tools on the Tamarind Bio platform, via its REST API or MCP server — no local GPUs required. Tamarind bundles popular open-source models for structure prediction (AlphaFold, Boltz, Chai, ESMFold), protein, binder, and de novo design (RFdiffusion, ProteinMPNN, BoltzGen), antibody and nanobody design and developability, protein-ligand docking (DiffDock, Autodock Vina), binding-affinity prediction, MSA generation, and molecular dynamics. Use when the user mentions Tamarind or tamarind.bio, wants to run any of these open-source tools in the cloud, references app.tamarind.bio/api or the x-api-key header, or needs to submit batches of sequences for structural or biophysical characterization.
+license: MIT
+compatibility: Requires Python 3.10+, a Tamarind Bio account, and an API key from app.tamarind.bio. Uses the `requests` library against the public REST API (no dedicated Python SDK exists). Network access required. Optional MCP server at mcp.tamarind.bio/mcp for agent hosts.
+metadata:
+  version: "1.1"
+  skill-author: Tamarind Bio
+  trigger-keywords: protein structure prediction, AlphaFold, Boltz, Chai, ESMFold, protein design, binder design, de novo design, antibody design, nanobody, protein-ligand docking, DiffDock, Autodock Vina, binding affinity, MSA generation, inverse folding, ProteinMPNN, RFdiffusion, BoltzGen, cloud GPU biology, structure prediction API, x-api-key, developability, adme, enzyme, peptide, protein language models, molecular design
+  openclaw:
+    primaryEnv: TAMARIND_API_KEY
+    envVars:
+    - name: TAMARIND_API_KEY
+      required: true
+      description: Tamarind Bio API key sent as the x-api-key header.
+---
+
+# Tamarind Bio
+
+Tamarind Bio is a cloud platform that runs computational biology tools — structure prediction, protein and antibody design, docking, binding-affinity, MSA generation, and molecular dynamics — on managed GPUs. Users submit sequences or structures and get back predicted structures, designs, and biophysical scores, without provisioning their own hardware. It exposes hundreds of tools (AlphaFold, Boltz-2, Chai-1, RFdiffusion, ProteinMPNN, BoltzGen, ESMFold2, DiffDock, Autodock Vina, and many more) through one uniform job API.
+
+**Official docs:** [app.tamarind.bio/api-docs](https://app.tamarind.bio/api-docs) · platform UI at [app.tamarind.bio](https://app.tamarind.bio)
+
+## Canonical sources — fetch these, don't rely on a stale copy
+
+Tamarind publishes live, machine-readable sources. Prefer fetching them at runtime over trusting any hardcoded list — tool names, schemas, and endpoints change frequently:
+
+- **`https://app.tamarind.bio/llms.txt`** — LLM index: links to the spec, API docs, and MCP guide.
+- **`https://app.tamarind.bio/openapi.yaml`** — OpenAPI 3.0 spec for the 8 core job endpoints (submit-job/-batch, jobs, result, upload, files, delete-job/-file; auth `ApiKeyAuth`). Fetch it for those exact shapes. Discovery/management endpoints (`/tools`, `/usage-statistics`, pipelines, …) aren't in it — use the MCP/REST discovery tools for those.
+- **`https://docs.tamarind.bio/llms.txt`** — documentation index; every page has a `.md` form (e.g. `docs.tamarind.bio/tamarind/batch.md`, `/tamarind/api.md`, `/tamarind/pipelines.md`).
+- **Live tool discovery** — `GET /tools` (REST) or MCP `getAvailableTools` + `getJobSchema(jobType)` are the source of truth for what tools exist and their parameters.
+
+This skill teaches the surface + the non-obvious behaviors those sources don't spell out (see the reference files). When in doubt about a shape, fetch `openapi.yaml`.
+
+## When to use this skill
+
+Use Tamarind when the user wants to:
+
+- **Predict structure** of a protein, complex, or protein-ligand system (AlphaFold, Boltz-2, Chai-1, ESMFold2, Chai/Boltz cofolding)
+- **Design proteins or binders** (RFdiffusion, BoltzGen, BindCraft, ProteinMPNN/LigandMPNN inverse folding)
+- **Design or characterize antibodies/nanobodies** (sequence generation, humanization, developability, immunogenicity)
+- **Dock small molecules** to a protein (DiffDock, Autodock Vina) or predict **binding affinity**
+- **Generate MSAs** for downstream folding
+- **Run molecular dynamics** or other biophysical workflows on managed GPUs
+- **Batch-screen** many sequences or designs through the same tool
+- **Chain tools** into pipelines (e.g. design → fold → score) using the output of one job as the input of the next
+
+This skill is the right fit when the work should run on Tamarind's managed cloud rather than on a local install. For purely local cheminformatics or one-off sequence I/O, use a local library (RDKit, BioPython) instead.
+
+## Access and authentication
+
+1. Sign in at [app.tamarind.bio](https://app.tamarind.bio) and create an API key from the account/API settings.
+2. Authenticate every REST request with the `x-api-key` header.
+3. **Never hardcode the key.** Read it from the `TAMARIND_API_KEY` environment variable or a `.env` file (use `python-dotenv`). Never commit keys to source control.
+
+**Pricing:** Every user gets **10 free jobs**. For larger usage, contact [info@tamarind.bio](mailto:info@tamarind.bio) to purchase a subscription.
+
+```bash
+export TAMARIND_API_KEY="your_api_key"
+# List available tools
+curl https://app.tamarind.bio/api/tools \
+  -H "x-api-key: $TAMARIND_API_KEY"
+```
+
+**Base URL:** `https://app.tamarind.bio/api/`
+
+There is **no official Python SDK** — the PyPI package named `tamarind` is an unrelated Neo4j tool. Do not `uv pip install tamarind`. Write plain `requests` calls against the REST API (the endpoint shapes are in `openapi.yaml`), or use the MCP server for agent hosts.
+
+## Two ways to call Tamarind
+
+### MCP server (best for AI agents)
+
+Tamarind hosts an MCP server at `https://mcp.tamarind.bio/mcp` (API-key auth via the `X-API-Key` header). When your agent host supports MCP, prefer it — the tools mirror the REST API with agent-friendly schemas:
+
+- `listModalities()` / `listTags()` — the live filter vocabulary (molecule type / function) with labels + tool counts; call these to learn valid `modality`/`function` values instead of hardcoding
+- `getAvailableTools(modality?, function?, search?, custom?)` — discover tools (`category`/`tag` are deprecated aliases still honored)
+- `getJobSchema(jobType)` — exact parameter schema for a tool, plus an `exampleJob` starting payload (validate it before submitting)
+- `validateJob(jobName, type, settings)` — dry-run validation before submitting
+- `submitJob(jobName, type, settings)` / `submitBatch(batchName, type, settings[], jobNames[])`
+- `getJobs(jobName?, batch?, limit?, includeSequences?)` — list/inspect jobs and statuses (the bulky per-job input blob is omitted by default; pass `includeSequences=true` to keep it)
+- `getJobLogs(jobName)` — fetch output logs for debugging
+- `listJobFiles(jobName)` — list output files (returns `s3Path` for chaining)
+- `getResult(jobName, fileName?)` — download results
+- `uploadFile(filename)` — presigned upload URL; or `uploadFileContent(filename, content, encoding?)` to send file content through MCP when the host can't reach S3 (sandboxed agents)
+
+Scope note: MCP query tools (`getJobs`, `getResult`, `listJobFiles`, …) are scoped to the authenticated account.
+
+### REST API (universal)
+
+Use plain HTTP with `requests` — the endpoint shapes are in `openapi.yaml`. The core loop is below; `references/workflows.md` has full recipes.
+
+## Core workflow
+
+Always follow discover → schema → validate → submit → poll → results. Do not hardcode tool names or settings — the catalog changes frequently.
+
+```python
+import os, time, requests
+
+BASE = "https://app.tamarind.bio/api"
+HEADERS = {"x-api-key": os.environ["TAMARIND_API_KEY"]}
+
+# 1. Discover tools. REST /tools returns the full list; filter client-side.
+tools = requests.get(f"{BASE}/tools", headers=HEADERS).json()
+alphafold = next(t for t in tools if t["name"] == "alphafold")
+
+# 2. Get the exact schema for the chosen tool.
+#    REST: each /tools entry already includes its inline `settings` schema
+#          (parameter list) — find the entry whose name == your job type.
+#    MCP:  getJobSchema(jobType) returns the same per-tool detail.
+
+# 3. Submit a job. `settings` is tool-specific — match the schema exactly.
+payload = {
+    "jobName": "my-alphafold-run",          # ^[a-zA-Z0-9_-]+$, <=100 chars, unique
+    "type": "alphafold",
+    "settings": {
+        "sequence": "MKTVRQERLKSIVRILERSKEPVSGAQLAEELSVSRQVIVQDIAYLRSLGYNIVATPRGYVLAGG",
+        "numRecycles": 3,
+    },
+}
+resp = requests.post(f"{BASE}/submit-job", headers=HEADERS, json=payload)
+resp.raise_for_status()   # 200 ok; 400 bad request; 403 budget exceeded; 401 unauthorized
+
+# 4. Poll for completion.
+#    NOTE the response shape: GET /jobs?jobName=<name> returns the job ROW
+#    directly (no "jobs" wrapper); the list query (no jobName) returns
+#    {"jobs": [...]}. Don't index ["jobs"][0] on the by-name response.
+while True:
+    job = requests.get(f"{BASE}/jobs", headers=HEADERS,
+                       params={"jobName": "my-alphafold-run"}).json()
+    if job["JobStatus"] in ("Complete", "Stopped", "Deleted"):
+        break
+    time.sleep(30)
+
+# 5. Retrieve results. POST /result returns a presigned URL *string*;
+#    GET that URL to download the actual results zip (two-step).
+url = requests.post(f"{BASE}/result", headers=HEADERS,
+                    json={"jobName": "my-alphafold-run"}).text.strip('"')
+open("my-alphafold-run.zip", "wb").write(requests.get(url).content)
+```
+
+For the agentic version of this loop using MCP tools, and for richer examples, see `references/workflows.md`.
+
+## Discovering tools
+
+The catalog has hundreds of tools. Always enumerate at runtime — never rely on a hardcoded list.
+
+**REST** `GET /tools` returns the **full list** (it does not filter server-side); each item is `{name, displayName, github, paper, description, settings}` where `settings` is that tool's inline parameter schema. Filter client-side:
+
+```python
+tools = requests.get(f"{BASE}/tools", headers=HEADERS).json()   # a list
+boltz = [t for t in tools if "boltz" in t["name"].lower()]
+```
+
+Note: both surfaces return one row per tool name — REST `/tools` and MCP `getAvailableTools` are both deduplicated (the MCP keeps the newest tool version), so a name match returns a single row.
+
+**MCP** `getAvailableTools(search=..., modality=..., function=...)` filters server-side and adds `categories`/`tags` per tool (`category`/`tag` are deprecated aliases of `modality`/`function`, still honored). Don't hardcode the vocabulary — it drifts. Get the live values from `listModalities()` / `listTags()` (each returns `value`, `label`, `description`, and `toolCount`), or read the `availableCategories` / `availableTags` facet arrays returned on every `getAvailableTools` response. Modalities are molecule types (protein, antibody, peptide, small-molecule, nucleic-acid, …); functions are what a tool does (structure-prediction, binder-design, protein-ligand-docking, …).
+
+A representative set of widely-used tools (verify with `/tools`): `alphafold`, `boltz` (Boltz-2), `chai` (Chai-1), `esmfold` / `esmfold2`, `rfdiffusion`, `proteinmpnn`, `ligandmpnn`, `boltzgen`, `bindcraft`, `diffdock`. See `references/tool_catalog.md` for the full category/tag map and how to read tool metadata.
+
+## Choosing the right tool
+
+The catalog has many tools per task; **don't hardcode a favorite — filter by `function` (and `modality`), then read each candidate's `description` and match it to the user's actual goal** (input you have, output you need, constraints like speed or "no MSA"). The `description` and `tags` fields are the public "what it's for" signal; let them, plus `validateJob`, drive the pick. Quick orientation by task:
+
+- **Fold a single protein / complex** (`function=structure-prediction`): the AlphaFold3-class reproductions — `boltz`/`chai`/`openfold`/`protenix`/`intfold` — are the accurate default for **everything**, including protein-only systems; they also handle **nucleic-acid + small-molecule complexes**, so reach for them whenever a ligand/RNA/DNA is part of the system (and `boltz` adds binding-affinity). `alphafold` (AF2) remains a solid choice for monomers + multimers (join chains with `:`). `esmfold` is single-sequence (no MSA) and fast — reach for it when you want speed and have no MSA; `esmfold2` is newer and conditions on an MSA by default (its `model` setting offers a faster single-sequence mode). Specialized folders exist for antibodies (`abodybuilder`, `immunebuilder`), cyclic peptides (`highfold`), and conformational ensembles (`afcluster`, `alphaflow`) — filter and read descriptions.
+- **Design a binder** (`function=binder-design`): `bindcraft` (de novo miniprotein binders) and `boltzgen` (binders for protein **and** small-molecule targets, incl. nanobodies/antibodies/peptides) are the go-to de novo binder tools; `rfdiffusion` also does binder design and is the pick for **motif scaffolding** / diversifying an existing backbone. Antibody-specific generators live under `function=antibody-design`.
+- **Design sequence for a known backbone** (`function=inverse-folding`): `proteinmpnn` (general), `ligandmpnn` (ligand-aware), plus thermostable/soluble/antibody MPNN variants. Inverse folding takes a **structure** and emits **sequences** — fold them back to verify (see chaining).
+- **Dock a small molecule** (`function=protein-ligand-docking`): prefer `boltz`/`chai` — they co-fold the ligand into the complex and predict the bound structure rather than docking into a fixed receptor; reach for `autodock-vina` when you need fast, large-scale screening against a known pocket.
+- **Predict binding affinity** (`function=binding-affinity`) or **generate an MSA** (search `msa`) — filter and read.
+
+When the user names a specific tool, evaluate that one **and** sanity-check the alternatives in its `tag` group — a faster or more appropriate sibling often exists. When unsure, `getJobSchema`/`validateJob` to confirm a candidate actually accepts the input you have before committing.
+
+## Job settings, schemas, and validation
+
+Each tool has its own `settings` schema. Fetch it before submitting:
+
+- **REST** `/tools` entry: each `settings` param is a **trimmed** dict. Only `name` and `required` are always present; `type`, `default`, `description`, `options` appear only when relevant (≈60% have `type`) — so use `param.get("type")`, not `param["type"]`. The advanced gating keys (`exclude`, `conditionals`) are **NOT in the REST response** at all.
+- **MCP** `getJobSchema(jobType)`: the **full** schema, including `exclude`, `conditionals`, and bounds. Use MCP when you need to reason about those gating keys. (`restrictOrgs` is stripped on both surfaces — an org-gated param you can't use is simply omitted; see `references/api_reference.md`.)
+
+**Always `validateJob` (MCP) before submitting** — it's the reliable guard. It runs the same validation as `/submit-job` without submitting, and surfaces the first missing/invalid field. Don't try to hand-derive which fields to strip from the schema keys (over REST you can't see them anyway) — let `validateJob` tell you. (The response may include a `source` field, e.g. `"static-fallback"` — an internal note on which schema source validated; `valid: true/false` is the signal you act on.)
+
+`validateJob` echoes a `normalized` view of your settings with defaults filled in. Submit the same clean `settings` you validated; treat `normalized` as informational (it can carry defaults you didn't set, and for some tools platform-managed fields), so build your submit from your own settings rather than the normalized blob.
+
+**Sequences:** amino-acid string; separate chains of a multimer with a colon (`:`), e.g. `"MVLS...:EVQL..."`. Note that some tools (e.g. `boltz`, `chai`) require more than `sequence` — `boltz` also requires `inputFormat` (and accepts `yamlFile`/`molecules`). Always `getJobSchema`/`validateJob` to learn a tool's required fields; don't assume `sequence` alone suffices.
+
+**Platform-internal fields** — never set these yourself; the platform owns them: `submit_method`, `monomer_msa`, `msa`. See `references/api_reference.md` for the full field-handling rules.
+
+**Surface consequential choices before submitting, don't default silently.** When the request fully specifies what to run, proceed. But when it's open-ended, or when a setting materially changes the results, runtime, or cost (model/variant, number of samples or seeds, MSA on/off, GPU tier, batch size), present the meaningful options plus the default you'd otherwise apply and let the user pick **before** you submit — rather than choosing silently and reporting it after the job is queued. `getJobSchema` and `validateJob`'s `normalized` show exactly which knobs you're filling in on the user's behalf, so you can flag the few worth a quick confirm. This matters most for **batches**, where one shared-settings choice multiplies across every job.
+
+## File inputs (PDB, CIF, SDF, …)
+
+Tools with file parameters accept input three ways:
+
+1. **Upload first, then reference by bare filename.** `PUT /upload/{filename}`, or MCP `uploadFile` → presigned URL → `curl -X PUT -T file "<url>"`. If your host can't reach S3 (a sandboxed agent with no outbound network), use MCP `uploadFileContent(filename, content, encoding?)` to send the file's content through the MCP channel instead — text by default, `encoding="base64"` for binary. The object lands at the S3 key `{email}/{filename}`, **but you reference it in `settings` by the bare `filename` only** (e.g. `"targetFile": "GLP1R_ECD.pdb"`) — the platform scopes it to your account automatically. **Do NOT prefix the email**: passing `{email}/{filename}` double-prefixes the lookup and `submit-job` 400s with `"The following files have not been uploaded: <email>/<file>"`. Confirm the exact name the store registered with MCP `getFiles(search=...)` / REST `GET /files` (a flat list of bare names).
+2. **Reference a prior job's output** by its path: `JobName/path/to/file.ext` (this is how you chain jobs — see below).
+3. **Inline content.** Send the file's text content directly as the field value.
+
+**Foot-gun:** for a file-typed parameter, a **plain string value is treated as inline file content**, not as a path to an existing object. To point at an already-uploaded file, use the bare `filename` (not the `{email}/...` S3 key) or, for a prior job's output, the `JobName/...` path form — not a bare string you expect to resolve to new content.
+
+**`validateJob` notes.** The response may carry a `source` field (e.g. `"static-fallback"`) — it labels how the tool's *schema* was resolved (built-in tools always report `static-fallback`), **not** whether the validator was reachable, so act on `valid`, not `source`. For file params: reference an uploaded file by its **bare filename** (above) — a bare name resolves to your account-scoped object, whereas an email-prefixed string can be read as inline content and fail the file-type check (`"... must contain ATOM records"`). And passing **inline** file content makes `validateJob` upload it synchronously before validating, which can be slow; prefer referencing an uploaded file by name (above). If a dry-run is slow, skip it and let `submit-job` validate.
+
+## Chaining jobs into pipelines
+
+A finished job's output becomes the next job's input — no download/re-upload. **Match the input type the next tool actually wants:** a sequence-design tool (ProteinMPNN) emits *sequences*, so you fold them by passing each as a `sequence`; a tool that takes a *file* parameter takes a path.
+
+The cleanest design→fold chain is the MCP `submitBatch(fromJob=...)`, which reads a completed design job's generated sequences and folds each as one job:
+
+```
+# ProteinMPNN designs sequences -> fold every one with AlphaFold, one call:
+submitBatch(batchName="verify-designs", type="alphafold", fromJob="my-proteinmpnn-job")
+```
+
+For a **file** input (e.g. a tool that takes a `.pdb`/`.cif`), reference a prior job's output by the path form `JobName/path/to/file.ext` in that file parameter. Two cautions, both confirmed by validation: (1) match the parameter's required **file type** — e.g. AlphaFold's `templateFiles` accepts only `.cif` and is a list, and is gated behind `templateMode: "custom"`; (2) `templateFiles` is for *structural templates*, not for "fold this designed sequence" — to fold a sequence, pass `sequence`. Always `getJobSchema`/`validateJob` to confirm a file param's type/conditions before chaining into it.
+
+To discover a job's exact output paths, use MCP `listJobFiles(job1)` — it returns each file's `s3Path`, usable directly in the next `submitJob`. (The REST `GET /files` lists your account's *uploaded* files as a flat name list; it does not enumerate a job's outputs.) Tamarind also supports saved **pipelines**: build one in the UI, then drive it with `/run-pipeline` (`{pipelineName, initialInputs, inputs}`) or define `stages[]` inline via `/submit-pipeline` (each stage names a `task` + `toolSettings`, using `"pdbFile": "pipe"` to thread one stage's output into the next). See `references/workflows.md`.
+
+## Batch submission
+
+Submit many jobs of the **same tool** in one call. The Python form uses parallel `settings[]` and `jobNames[]` arrays (same length, up to 100):
+
+```python
+requests.post(f"{BASE}/submit-batch", headers=HEADERS, json={
+    "batchName": "egfr-binder-screen",
+    "type": "alphafold",
+    "jobNames": ["seq1", "seq2", "seq3"],
+    "settings": [{"sequence": "..."}, {"sequence": "..."}, {"sequence": "..."}],
+    # optional: "maxRuntimeSeconds": 3600, "weightedHoursBudget": 100,
+    # (some accounts also accept an optional "gpuType" — confirm with support)
+})
+```
+
+**Poll the batch *parent* on `batchStatus`, not subjob `JobStatus`.** A batch creates a parent job (`Type: "batch"`) plus subjobs. Subjobs flip to `Complete` as soon as they finish computing, but the batch then spends a few minutes **aggregating** results into the final downloadable output. Fetch the parent by name and watch `batchStatus`:
+
+```python
+import time
+while True:
+    # ?jobName= returns the parent ROW directly (no "jobs" wrapper)
+    parent = requests.get(f"{BASE}/jobs", headers=HEADERS,
+                          params={"jobName": "egfr-binder-screen"}).json()
+    bs = parent.get("batchStatus")
+    if bs == "Complete":
+        break
+    if bs in ("Stopped", "AggregationFailed"):
+        raise RuntimeError(parent.get("AggregationError", bs))
+    time.sleep(15)   # Running / Aggregating -> keep waiting
+# When Complete, the parent carries a presigned `resultUrl` and a `statuses`
+# subjob tally ({Complete, Running, In Queue, Stopped}).
+open("batch.zip", "wb").write(requests.get(parent["resultUrl"]).content)
+```
+
+Add `includeSubjobs=true` to `GET /jobs?batch=<name>` to list per-subjob rows.
+
+## Job status lifecycle
+
+Single jobs report `JobStatus`; batch parents report `batchStatus` (poll that for batches — see above).
+
+| Status | Meaning |
+|---|---|
+| `In Queue` | Accepted, waiting for capacity |
+| `Running` | Executing on a worker |
+| `Complete` | Finished successfully — results available |
+| `Stopped` | Stopped (failure, timeout, manual stop, or budget) |
+| `Deleted` | Job was deleted out-of-band |
+| `Aggregating` | (batch parent only) subjobs done; building the final output |
+| `AggregationFailed` | (batch parent only) aggregation step failed |
+
+Completed jobs carry a `Score` (tool-specific metrics, e.g. pLDDT/pTM/ipTM for folding) and `WeightedHours`. Treat `Complete`/`Stopped`/`Deleted` (and `AggregationFailed` for batches) as terminal; poll on a 15-30s interval. **Break your poll loop on any terminal status, not just `Complete`/`Stopped`** — a job that goes `Deleted` mid-poll would otherwise loop forever. For a `Stopped` job, fetch `getJobLogs(jobName)` to see why. `WeightedHours` is the usage unit billed per job; cap a batch with `weightedHoursBudget`, and a `403` on submit means a budget was hit (see `references/api_reference.md` and the `/usage-statistics` endpoint).
+
+## Error handling
+
+| Code | Meaning | Action |
+|---|---|---|
+| 400 | Bad request / invalid settings | Re-check against the schema; run `validateJob` first |
+| 401 | Unauthorized | Check `x-api-key` |
+| 403 | Budget exceeded (org/team) | Lower scope or raise the budget |
+| 429 | Rate limited | Back off and retry |
+| 500 | Server error | Retry; if persistent, contact support |
+
+## Reference files
+
+The `openapi.yaml` spec is the source of truth for endpoint shapes; these files add the behaviors and gotchas the spec doesn't spell out:
+
+- `references/examples.md` — **validated** `settings` payloads per common tool (alphafold/boltz/diffdock/autodock-vina/proteinmpnn/batch), a copy-paste self-check, the "what fails and the exact error" list, and output-shape notes. Start here for a working payload.
+- `references/api_reference.md` — endpoint quick-reference + the non-obvious shapes: `/jobs` by-name returns a bare row (not `{jobs:[...]}`), `/result` is a two-step download, batch parents poll on `batchStatus`, `/files` is a flat name list, the `settings` field-handling rules.
+- `references/tool_catalog.md` — category/tag map, how to read tool + parameter metadata, common tool families.
+- `references/workflows.md` — end-to-end recipes: fold a sequence, validate-before-submit, upload + reference a file, design→fold chaining, batch screen with aggregation polling, usage stats, pagination, and the non-blocking submit-now/check-later pattern for long jobs.

package/skills/tamarind/references/api_reference.md

@@ -0,0 +1,165 @@
+# Tamarind Bio REST API reference
+
+**Spec:** the OpenAPI spec at `https://app.tamarind.bio/openapi.yaml` (3.0, auth `ApiKeyAuth`) covers the 8 **core job endpoints** (`/submit-job`, `/submit-batch`, `/jobs`, `/result`, `/upload/{filename}`, `/files`, `/delete-job`, `/delete-file`) — fetch it for those exact shapes. It does **not** include the discovery/management endpoints (`/tools`, `/usage-statistics`, `/submit-pipeline`, `/run-pipeline`, `/stop-job`) — for those, use this file + the live MCP `getAvailableTools`/`getJobSchema`/`getJobs`. This file also adds the behaviors no spec spells out (response-shape-by-query, two-step result download, batch aggregation polling, REST-vs-MCP field differences).
+
+Base URL: `https://app.tamarind.bio/api/`
+Authentication: `x-api-key: <YOUR_KEY>` header on every request.
+Interactive docs: [app.tamarind.bio/api-docs](https://app.tamarind.bio/api-docs) · markdown docs at [docs.tamarind.bio](https://docs.tamarind.bio)
+
+There is no official Python SDK. Call the API with `requests` (Python) or `curl`. An MCP server (`https://mcp.tamarind.bio/mcp`, `X-API-Key` header) exposes the same operations with agent-friendly schemas.
+
+## Endpoints
+
+| Method | Path | Purpose |
+|---|---|---|
+| GET | `/tools` | List available tools and their inline parameter schemas. Returns the **full list** (no server-side filtering — filter client-side). |
+| POST | `/submit-job` | Submit one job. Body: `jobName`, `type`, `settings` (+ optional `projectTag`). |
+| POST | `/submit-batch` | Submit many jobs of the same tool. See payload shapes below. |
+| GET | `/jobs` | List/inspect jobs. Query: `jobName`, `batch`, `limit`, `startKey`, `organization`, `includeSubjobs`, `jobEmail`. |
+| POST | `/result` | Get a presigned download URL for job results (two-step — see below). Body: `jobName` (+ optional `fileName`, `pdbsOnly`, `jobEmail`). |
+| POST | `/stop-job` | Stop a running or queued job. Body: `jobName`. |
+| DELETE | `/delete-job` | Delete a job and its data. Body: `jobName`. |
+| PUT | `/upload/{filename}` | Upload a file (`--data-binary`; add `?folder=` to file it). Or get a presigned URL via MCP `uploadFile`. |
+| GET | `/files` | List your account's uploaded files as a flat array of filename strings. Query: `folder`, `includeFolders=true`. Does **not** enumerate a specific job's outputs — use MCP `listJobFiles` for that. |
+| DELETE | `/delete-file` | Remove a file/folder. Query: `filePath` or `folder`. |
+| POST | `/submit-pipeline` | Run a multi-step pipeline defined inline via `stages[]`. |
+| POST | `/run-pipeline` | Run a pipeline saved in the UI. Body: `pipelineName`, `initialInputs`/`inputs`. |
+| GET | `/usage-statistics` | Usage/billing. Query: `statistic` (`weighted_hours`/`jobs`), `scope` (`user`/org). |
+
+## Request shapes
+
+### GET /tools
+
+Returns a JSON **array**. Each element:
+
+```json
+{
+  "name": "alphafold",
+  "displayName": "AlphaFold",
+  "description": "Accurate and quick protein structure prediction ...",
+  "github": "https://github.com/...",
+  "paper": "https://...",
+  "settings": [ { "name": "sequence", "type": "sequence", "required": true, "description": "..." }, ... ]
+}
+```
+
+In each `settings` param, only `name` and `required` are guaranteed; `type`, `default`, `description`, `options` are present only when applicable (about 60% of params carry `type`). Read them with `param.get("type")`, not `param["type"]`.
+
+`settings` is the tool's inline parameter schema — read it directly, no separate schema endpoint over REST. The REST list is not filtered by query params; filter client-side on `name`/`displayName`/`description`. (The MCP `getAvailableTools` wraps the list as `{"totalTools", "tools":[...]}` and adds `categories`/`tags` per tool plus server-side `search`/`category`/`tag` filtering.)
+
+### POST /submit-job
+
+```json
+{
+  "jobName": "my-protein-analysis",
+  "type": "alphafold",
+  "settings": { "sequence": "MKT...", "numRecycles": 3 },
+  "projectTag": "proj_xxxxxxxx"
+}
+```
+
+- `jobName` — unique, `^[a-zA-Z0-9_-]+$`, 1-100 chars.
+- `type` — a tool name from `/tools`. The list changes often; never hardcode.
+- `settings` — tool-specific; match the schema from `/tools` (or MCP `getJobSchema`).
+- `projectTag` — optional `proj_...` ProjectId to file the job under a project.
+
+Response (200): a confirmation string like `myJobName submitted to queue.`
+
+### POST /submit-batch
+
+Two payload shapes appear in the official docs — the **Python** form uses parallel arrays; the **curl** form uses a `jobs[]` array of objects with a `tool` key. The parallel-array form matches the MCP `submitBatch` and is the recommended one:
+
+```json
+{
+  "batchName": "egfr-screen",
+  "type": "alphafold",
+  "jobNames": ["seq1", "seq2"],
+  "settings": [{ "sequence": "..." }, { "sequence": "..." }],
+  "maxRuntimeSeconds": 3600,
+  "weightedHoursBudget": 100
+}
+```
+
+curl-form alternative (same endpoint): `{ "tool": "<type>", "batchName": ..., "jobs": [{ "jobName": ..., "settings": {...} }, ...] }`.
+
+- `jobNames` and `settings` are parallel arrays, same length, 1-100 items, all using the same tool.
+- `maxRuntimeSeconds` — optional per-job timeout. `weightedHoursBudget` — optional budget cap.
+- The MCP `submitBatch` schema exposes `maxRuntimeSeconds` + `weightedHoursBudget`. Some accounts/tools may accept an optional `gpuType` (seen in the docs UI), but it isn't in `openapi.yaml` or the MCP schema — treat it as unverified and confirm with support before relying on it.
+
+### GET /jobs
+
+**Response shape depends on the query:**
+- **List / batch query** (no `jobName`, or `?batch=`/`?organization=`) → `{ "jobs": [...], "startKey": "...", "statuses": {...} }`.
+- **By-name** (`?jobName=<name>`) → the **job row object directly** (no `jobs` wrapper). Don't index `["jobs"][0]` on this response.
+
+Each job row includes `JobName`, `Type`, `JobStatus`, `Created`, `Started`, `Completed`, `Settings` (JSON string), `Score` (JSON string, tool metrics), `WeightedHours`. Use `startKey` for pagination past the `limit` (default 1000). Only top-level jobs return by default; add `includeSubjobs=true` for batch subjobs.
+
+**Batch parent rows** have `Type: "batch"` and carry `batchStatus`. Fetched by name (`?jobName=<batchName>`), a complete batch parent also includes `resultUrl` (presigned download). `batchStatus` transitions: `Running` → `Aggregating` → `Complete` (or `AggregationFailed`, with `AggregationError`). Poll the parent's `batchStatus`, not subjob `JobStatus` — subjobs go `Complete` before the aggregated output is ready.
+
+**Discriminate batch vs single by `Type == "batch"` (or presence of `batchStatus`), not by `statuses`.** A by-name response can carry a `statuses` tally even for a single (non-batch) job, so `statuses` presence is not a reliable batch signal.
+
+### POST /result (two-step download)
+
+POST returns a presigned URL as a **bare string** (not JSON). Fetch that URL with a second GET to download the results zip:
+
+```python
+url = requests.post(f"{BASE}/result", headers=H, json={"jobName": "myJob"}).text.strip('"')
+open("myJob.zip", "wb").write(requests.get(url).content)
+```
+
+Optional body fields: `fileName` (one file instead of the zip), `pdbsOnly: true` (PDB outputs only), `jobEmail` (a teammate's job, if permitted).
+
+## Status codes
+
+| Code | Meaning |
+|---|---|
+| 200 | Success |
+| 400 | Bad request — invalid parameters/settings |
+| 401 | Unauthorized — invalid/missing `x-api-key` |
+| 403 | Budget exceeded (org/team) |
+| 429 | Rate limited |
+| 404 | Not found (e.g. unknown job) |
+| 500 | Server error |
+
+## Field-handling rules (important)
+
+**The REST and MCP schemas expose different fields.** The REST `/tools` entry
+gives a trimmed per-param view — `{name, type, required, default, description, options}`.
+The advanced gating keys `exclude` and `conditionals` appear **only in MCP
+`getJobSchema`**, not in REST `/tools` (`restrictOrgs` is no longer returned by
+either surface — see below). So don't try to hand-derive what to strip from REST
+schema keys — they aren't there. The reliable guard on
+both surfaces is **`validateJob`** (MCP): it runs `/submit-job`'s exact validation
+without submitting and returns the first error.
+
+- **Build your submit from your own settings, not `validateJob`'s `normalized` output.**
+  `normalized` is informational (defaults filled in, sometimes platform-managed
+  fields). Submit the same clean settings you validated, not the normalized echo.
+- **Platform-internal routing fields** — `submit_method`, `monomer_msa`, `msa` are
+  set by the platform. Never pass them.
+- **`restrictOrgs`** — org-gated parameters. `getJobSchema` no longer returns this
+  key (it's stripped server-side): a parameter your account isn't authorized for is
+  dropped from the schema entirely, and any param you do see is one you may set. So
+  you won't encounter `restrictOrgs` in a response — don't look for it.
+- **`conditionals`** (MCP schema only) — a field only applies when another field
+  has a given value (e.g. `pairMode` applies only when `useMSA` is `true`). Don't
+  send conditioned fields when their condition isn't met.
+- **`exclude: [...]`** (MCP schema only) — marks a field as UI/pipeline-only for a
+  surface. Treat it as advisory; `validateJob` is the authority on what a given
+  submission accepts.
+- **`required: true`** — must be present. Some tools require more than `sequence`
+  (e.g. `boltz` requires `inputFormat`). Run `validateJob` to get the first
+  missing/invalid field before submitting.
+- **File-typed fields with a plain string value are treated as INLINE CONTENT**,
+  not a path. To reference an **uploaded file**, use its **bare filename**
+  (`target.pdb`) — the platform scopes it to your account, so do NOT email-prefix
+  it. The `{email}/{filename}` form is the underlying S3 key, and passing it makes
+  `submit-job` 400 with `"The following files have not been uploaded: <email>/<file>"`.
+  To reference a **prior job's output**, use `JobName/path/to/file.ext`. Confirm the
+  exact registered name with `getFiles` / `GET /files` (a flat list of bare names).
+
+## Authentication and secrets
+
+- Read the key from `TAMARIND_API_KEY` (env or `.env`); never hardcode or commit it.
+- The same key authenticates REST (`x-api-key`) and the MCP server (`X-API-Key`).
+- Query operations are scoped to the authenticated account (and, with `organization=true`/`jobEmail`, to your org if permitted).

package/skills/tamarind/references/examples.md

@@ -0,0 +1,132 @@
+# Tamarind Bio — validated examples & output shapes
+
+**The freshest example for any tool is the `exampleJob` field MCP `getJobSchema(<tool>)`
+now returns** — an `{jobName, type, settings}` built from each param's example/default
+(with an `exampleJobNote`; org-gated params you can't use are omitted, file params get
+placeholder filenames). It's the best starting point, but **run `validateJob` on it
+before submitting** — it's assembled from per-param examples, not a guaranteed-valid
+payload, so a given tool's `exampleJob` can need a tweak. The payloads below are a
+`validateJob`-confirmed fallback for REST callers or when you want a worked example.
+Tool schemas evolve — if one stops validating, re-fetch with `getJobSchema(<tool>)` /
+`GET /tools`. Sequences here are illustrative; swap your own.
+
+**File params (`proteinFile`, `pdbFile`, `ligandFile`, …) need a real file value** —
+either the **bare filename** of an uploaded file (`target.pdb` — NOT email-prefixed),
+a prior-job output **path** (`JobName/out/x.pdb`), or
+**inline PDB/SDF-format text** (multi-line `ATOM`/`HETATM` records). The
+`<...>` placeholders below are NOT valid as written — replace them. **Do not put an
+amino-acid sequence in a file param** — `validateJob` rejects it with
+`File ... must be of types: ["pdb"]`. (A sequence goes in `sequence`, a structure
+goes in a file param.)
+
+`BASE = "https://app.tamarind.bio/api"`, `HEADERS = {"x-api-key": <key>}`.
+
+## Self-check (run this first to confirm the skill works for you)
+
+Read-only + dry-run, no submission, no cost. Confirms the discover → schema →
+validate loop end-to-end:
+
+```python
+import os, requests
+BASE, HEADERS = "https://app.tamarind.bio/api", {"x-api-key": os.environ["TAMARIND_API_KEY"]}
+
+# 1. discovery reachable?
+tools = requests.get(f"{BASE}/tools", headers=HEADERS).json()
+assert isinstance(tools, list) and any(t["name"] == "alphafold" for t in tools), "tools endpoint"
+
+# 2. validate a known-good payload (MCP validateJob; or skip if REST-only)
+#    expect {"valid": true, ...}
+```
+
+With the MCP server: `validateJob(jobName="selfcheck", type="alphafold",
+settings={"sequence": "MKTAYIAKQRQISFVKSHFSRQLEERLGLIE"})` → `valid: true`.
+
+## Validated input payloads
+
+### AlphaFold — monomer
+```json
+{ "sequence": "MKTAYIAKQRQISFVKSHFSRQLEERLGLIEVQAPILSRVGDGTQDNLSGAEKAVQVKVKALPDAQFEVVHSLAKWKR",
+  "numModels": "1", "numRecycles": 3 }
+```
+Only `sequence` is required; everything else has a default. `numModels` is a string
+dropdown (`"1"`–`"5"`).
+
+### AlphaFold — multimer (colon-separated chains)
+```json
+{ "sequence": "MKTAYIAKQRQISFVKSHFSRQLEERLGLIE:DIQMTQSPSSLSASVGDRVTITCRASQSISSYLN" }
+```
+Join chains with `:`. No separate "multimer" flag — chain count drives it.
+
+### Boltz-2 — sequence mode
+```json
+{ "inputFormat": "sequence",
+  "sequence": "MKTAYIAKQRQISFVKSHFSRQLEERLGLIEVQAPILSRVGDGTQDNLSGAEKAVQVKVKALP" }
+```
+`inputFormat` is **required** (`"sequence"` / `"list"` / `"molecules"` / `"yaml"`).
+Omitting it fails — see "What fails" below.
+
+### DiffDock — protein + SMILES ligand
+```json
+{ "ligandFormat": "SMILES",
+  "ligandSmiles": "CC(=O)Oc1ccccc1C(=O)O",
+  "proteinFile": "<uploaded-path-or-inline-PDB-text>" }
+```
+`ligandFormat` chooses the conditional field: `"SMILES"` → `ligandSmiles`;
+`"sdf/mol2 file"` → `ligandFile`. `proteinFile` is a file param — pass an uploaded
+file's bare filename (`target.pdb`, not email-prefixed), a prior-job path
+(`JobName/...`), or inline PDB text (see file-input rules in `api_reference.md`).
+
+### Autodock Vina — protein + SMILES ligand (classical docking into a pocket)
+```json
+{ "receptorFile": "receptor.pdb",
+  "ligandFormat": "smiles",
+  "ligandSmiles": "CC(=O)Oc1ccccc1C(=O)O",
+  "boxX": 15.19, "boxY": 53.903, "boxZ": 16.917,
+  "width": 20, "height": 20, "depth": 20 }
+```
+Unlike DiffDock, Autodock Vina docks into a **fixed pocket**, so it requires a bounding
+box (`boxX/Y/Z` center + `width/height/depth`, all required) and the receptor in
+`receptorFile` (not `proteinFile`). Its `ligandFormat` enum is **lowercase**
+(`"smiles"` / `"sdf"`) — different from DiffDock's `"SMILES"` / `"sdf/mol2 file"`, so
+don't copy DiffDock's value across. `exhaustiveness` (default 8) is optional. `validateJob`-confirmed.
+
+### ProteinMPNN — design residues on a backbone
+```json
+{ "pdbFile": "<uploaded-path-or-inline-PDB-text>",
+  "designedResidues": { "A": "1 2 3 4 5" },
+  "numSequences": 4, "modelType": "proteinmpnn" }
+```
+Requires `pdbFile` + `designedResidues` (per-chain, space-separated resnums).
+`modelType` ∈ `proteinmpnn`/`ligandmpnn`/`solublempnn`/`hypermpnn`/`abmpnn`.
+Note `designedChains` is `exclude:["api"]` — don't send it over the API.
+
+### Batch (same tool, many jobs)
+```json
+{ "batchName": "screen-1", "type": "alphafold",
+  "jobNames": ["s1", "s2"],
+  "settings": [ { "sequence": "MKT..." }, { "sequence": "AVF..." } ] }
+```
+
+## What fails (and the exact error) — confirmed live
+
+- **Boltz without `inputFormat`** → `valid:false`, `Missing required boltz field "inputFormat"`. Always check required fields with `getJobSchema` first; `sequence` alone is not enough for boltz/chai.
+- **Building a submit from `validateJob`'s `normalized` blob** — `normalized` is informational (defaults filled in, sometimes platform-managed fields). Submit the clean `settings` you validated, not the normalized echo.
+- **File param given a bare string that isn't a real path** → treated as INLINE file content (uploaded as `<email>/<jobname>-<param>.<ext>`), not a reference. To point at an existing uploaded file use its **bare filename** (`target.pdb` — do NOT email-prefix it; `{email}/{filename}` is the S3 key and 400s as not-uploaded), or `JobName/...` for a prior job's output. Referencing a path that doesn't exist → `File ... has not been uploaded`.
+
+## Output shapes (describe, don't expect exact values)
+
+Outputs are non-deterministic (seed/model/MSA) — reason about the *shape*, not
+golden numbers.
+
+- **Job row `Score`** (JSON string on completed jobs): tool-family dependent.
+  - Folding (alphafold/boltz/chai/esmfold): `plddt`, `ptm`, and for complexes
+    `iptm` plus interface metrics (`ipSAE_*`, `pDockQ_*`). Higher pLDDT/pTM = more
+    confident; iptm/ipSAE gauge interface quality.
+  - Other families carry their own metrics — read the keys, don't assume.
+- **Results zip** (`POST /result` → presigned URL → GET): per-tool, typically the
+  structure files (`rank_*.pdb` / `*.cif`), a scores CSV, and logs. Use
+  `listJobFiles(jobName)` (MCP) to enumerate exact filenames before downloading.
+- **`WeightedHours`** on the row is the billing unit (see `usage-statistics`).
+
+To learn a specific tool's exact outputs, run one small job and `listJobFiles` it —
+don't hardcode filenames, which vary by tool and version.