Explore Apps

Paste an SMS flow or campaign send. Your browser counts every message the way carriers bill it (GSM-7 or UCS-2, merge tags filled), finds the characters that double the cost, and checks sender ID, opt-out, the opt-in confirmation, public shorteners, restricted content, quiet hours in every time zone and frequency, free. A paid review rewrites what needs it, or a paid draft writes a flow from a brief, and every text it writes is measured and checked again. Derived from the agent skill @coreyhaines31/sms (coreyhaines31/marketingskills, MIT).

Screen a giveaway's entrant list for duplicate, excluded and late entries, run a seeded SHA-256 draw anyone can re-check, and get the winner announcement and verification note written from the draw record. Derived from @composiohq/raffle-winner-picker (ComposioHQ/awesome-claude-skills).

Paste one attack tree (an indented AND/OR outline, or the attack-tree-construction skill's JSON export) and your browser rolls cost, difficulty, detection and time up the gates, lists every attack scenario, finds the cheapest, easiest, stealthiest and fastest one, the minimal cut sets and what your in-place controls leave open, free. Or describe a system and have a scored tree drafted. A review then checks gates and estimates, adds missing branches and writes a defence plan and a stakeholder brief, each checked against the analysis. Built on the agent skill @wshobson/attack-tree-construction (wshobson/agents, MIT).

Local character AI that runs entirely in your browser - no install, no account. Gemma 4 E2B writes the replies, DreamShaper draws the pictures, Kokoro speaks and Whisper listens, all on your device. Nothing you type is sent anywhere.

Paste one balance-sheet account's GL detail for the period with its beginning balance and its ending balance per GL. Your browser sorts every posting into additions, accruals booked, reversals of prior accruals, payments and settlements, reclasses and adjustments, FX translation or unclassified, foots the roll-forward, ties each line to the rows behind it and works any gap instead of plugging it: a close-package beginning balance that differs from the GL, postings dated outside the period and exact duplicates are tested alone and together, and only a combination that closes the gap exactly is called its cause. Last period's accruals are matched to this period's reversals. All free, before sign-in. Then a review decides whether the schedule can go into the close package, explains every line and exception with an owner, places unclassified rows and writes the support requests and a file note, and every amount it writes is checked against the schedule. Derived from the @anthropics/roll-forward skill (anthropics/financial-services-plugins, Apache-2.0).

Paste a Flutter app_en.arb and your browser checks it with a port of gen-l10n's own ICU parser and placeholder rules, and shows what every plural prints for each count in each language - free. A paid run annotates the template for translators or translates it into a new locale (or fills a locale's missing keys), and every returned message is re-checked. Derived from @flutter/flutter-localizing-apps (flutter/agent-plugins, BSD-3-Clause).

Paste an llms.txt and your browser checks it against the llmstxt.org v2 format and the proposal's own reference parser, line by line, and compares its links with your sitemap, URL list or repository tree - free, nothing uploaded, with a starter draft. A paid run writes a new llms.txt from your page list, or brings an existing one up to date for what changed, answering every finding; every returned link is checked against your input and the file is checked again. Derived from the agent skills @github/create-llms and @github/update-llms (github/awesome-copilot, MIT).

Drop a CSV or TSV, name its split, subject, group and time columns, and your browser runs the exploratory-data-analysis skill's own profile, missingness/split-leakage audit and outlier sensitivity, same output as the Python, free, rows never leave the page. Paid lanes review the split (resplit plan + scikit-learn code) or draft the EDA report from aggregates only. Derived from @k-dense-ai/exploratory-data-analysis (k-dense-ai/scientific-agent-skills, MIT).

Check a deepTools script for ChIP-seq, ATAC-seq or RNA-seq coverage tracks, QC and heatmaps before you run it. Free in the browser: an emulation of deepTools 3.5.6's own argument parser for all 20 commands, the argument-only exits in each tool, best-practice checks (RPGC without a genome size, extended RNA-seq reads, another assembly's genome size, missing Tn5 shift), the skill's workflow templates, and what each edit resolved. Paid lanes review the script with a corrected version and a line diff, or draft a pipeline from a description - both re-checked in the page. Derived from the agent skill @k-dense-ai/deeptools (k-dense-ai/scientific-agent-skills, MIT). Independent tool, not affiliated with or endorsed by the deepTools project or the Max Planck Institute of Immunobiology and Epigenetics.

Drop the JSON records of an ISO 13485, ISO 14971, ISO/IEC 17025 or ISO 15189 readiness pack - scope intake, document register, CAPA, traceability, supplier controls, QMSR transition, evidence manifest. Your browser runs the iso-standards-readiness agent skill's own structural checks free, with the same report as its Python scripts; a paid run writes a draft evidence review for authorized human assessment or drafts a missing controlled procedure. Derived from @k-dense-ai/iso-standards-readiness and @k-dense-ai/iso-13485-certification (k-dense-ai/scientific-agent-skills, MIT).

Check a whole-slide image manifest and plan PathML 3.0.5 tiling and inference before a computational-pathology run: the pathml skill's slide_manifest.py, plan_pipeline.py and plan_inference.py run free in your browser, then an AI plan review or a methods and provenance draft from counts only. Derived from @k-dense-ai/pathml (k-dense-ai/scientific-agent-skills, MIT). Independent tool, not affiliated with the PathML project or Dana-Farber. Research use only.

Paste or drop the dxapp.json of a DNAnexus app or applet. Your browser runs the skill's own validate_dxapp.py (same report and exit status as the Python, free); a paid run reviews it for dx build or writes a corrected manifest the page re-validates. Derived from @k-dense-ai/dnanexus-integration (k-dense-ai/scientific-agent-skills, MIT).

Paste a research brainstorm's idea register, criteria and scoring sheet: your browser runs the scientific-brainstorming skill's own register validator and weighted matrix (score intervals, one-at-a-time weight sensitivity, fragile ranks, decision left null), byte-identical to the Python, free. A paid run drafts the three files from session notes or runs the skill's adversarial review of the shortlist. Derived from @k-dense-ai/scientific-brainstorming (k-dense-ai/scientific-agent-skills).

Bulk RNA-seq differential expression in the browser: load a counts matrix and a sample sheet and the pydeseq2 0.5.4 pipeline runs free (size factors, dispersions, Wald tests, Cook's and independent filtering, BH-adjusted p-values, LFC shrinkage), then a review of what the result supports or a pydeseq2 script that reproduces it. Derived from the agent skill @k-dense-ai/pydeseq2 (k-dense-ai/scientific-agent-skills).

Paste a DESIGN.md written to Google's alpha spec: the official @google/design.md linter runs in your browser (broken references, WCAG contrast, section order, misspelled keys) with byte-identical CSS, Tailwind and DTCG exports, free. A paid run reviews what a coding agent would get wrong, or returns a repaired file the browser re-lints and diffs. Derived from the @nousresearch/design-md skill.

Paste a multi-model PDB MD trajectory: your browser aligns it and computes RMSD, per-residue RMSF, radius of gyration and ligand contacts exactly as MDAnalysis 2.9 does, and flags periodic wraps, drift and a ligand that left - free. A metered run reads what the frames support or writes the MDAnalysis script that reproduces them. Derived from the agent skill @k-dense-ai/molecular-dynamics (k-dense-ai/scientific-agent-skills).

Paste a lab-animal welfare cohort and get RELSA severity scores and KDE severity zones computed in your browser exactly as the relsa-severity-assessment skill's scripts do (free), then a reading against the reporting checklist or a Python script that reproduces them. Derived from @k-dense-ai/relsa-severity-assessment (k-dense-ai/scientific-agent-skills, K-Dense Inc.).

Drop a PIV image pair: your browser runs the openpiv 0.25.4 single-pass pipeline (cross-correlation, sub-pixel peaks, validation, outlier replacement, scaling, vorticity) free, then a paid run reviews what the vector field supports or writes an openpiv script that reproduces it. Derived from the agent skill @k-dense-ai/openpiv (k-dense-ai/scientific-agent-skills).

Set up a multi-server queue - servers, waiting room, arrival and service times, horizon, warm-up, seed. Your browser runs the SimPy skill's replication study exactly (same seeds, same numbers) with Student-t intervals, an M/M/c/K benchmark and a paired what-if, free; a paid run reads what the study supports or writes the SimPy script that reproduces it. Derived from @k-dense-ai/simpy (k-dense-ai/scientific-agent-skills).

Aim, charge and fire around planets in a turn-based space duel with five CPU levels. Source: https://github.com/linkingtechnologies/libre-arcade/tree/main/highmoon (GPL-3.0-or-later).

Paste computed entries (formulas with total energies): your browser builds the phase diagram exactly as pymatgen's PhaseDiagram does - formation energies, convex hull, energy above hull, decompositions - free, and flags mixed methods and ties. A paid run interprets the hull or writes the pymatgen script that reproduces it. Derived from the agent skill @k-dense-ai/pymatgen (k-dense-ai/scientific-agent-skills, K-Dense Inc.).

Paste a data table and a patsy formula: your browser fits OLS exactly as statsmodels does (classical and HC3 errors, R-squared, F, AIC, Breusch-Pagan, Jarque-Bera, Durbin-Watson, VIF, Cook's distance), free. A paid run explains what the model supports or writes the statsmodels script that reproduces and fixes it. Derived from @k-dense-ai/statsmodels (k-dense-ai/scientific-agent-skills, K-Dense Inc.).

Paste a small metabolic model (cobrapy reaction strings or a cobrapy JSON model). Your browser runs flux balance analysis exactly as cobrapy does - optimum, pFBA fluxes, flux variability, blocked reactions, gene and reaction knockouts, mass balance, the default-bound check - free. A paid run reads what the result supports or writes the cobrapy script that reproduces every number. Derived from the agent skill @k-dense-ai/cobrapy (k-dense-ai/scientific-agent-skills, K-Dense Inc.).

Paste two BED region sets - peaks and promoters, two replicates, two conditions. Your browser computes their overlap exactly as gtars 0.9.2 does (Jaccard, coverage both ways, overlap coefficient, intervals with a hit, nearest gaps) plus a seeded within-chromosome shuffle test and a universe Fisher test, free. A paid run reads what the overlap supports or writes the gtars script that reproduces every number. Derived from the agent skill @k-dense-ai/gtars (k-dense-ai/scientific-agent-skills, K-Dense Inc.).