Skills Catalog

When a NewsCoin drops from a verified poster matching your filters (tags, creator wallet age, max market cap, headline keywords), auto-buy with a preset NEAR budget. You wake up to a position, not a chart.

Stake-aware tiered take-profit / stop-loss ladders for any NewsCoin position. Sells in 25/25/25/25 chunks across configured price marks; trails the bottom mark on a runner.

Scores a fresh NewsCoin 0–100 on virality (tag heat, creator track record, headline punch, source quality). One-shot when you're staring at the launch and need to decide in 30 seconds.

Preview slippage and final fill price for a buy or sell of size N on a given coin's bonding curve, including segment transitions. Saves you from eating 6% slippage on a chunky entry.

Watch a list of high-conviction creators; alert via TG and DM the moment one of them launches a new coin, before it hits the trending feed.

Dollar-cost-average into a coin over N hours/days at fixed intervals, with optional volatility-based pause if price spikes >X% in any interval.

Daily scan of your NewsCoin holdings; flags coins entering decline (volume <30% of 7d avg, holders dropping, expiring soon). Tells you which to exit, not which to hold.

Rebalance your NewsCoin bag back to target weights on a schedule. Sells winners, tops up laggards, respects min-trade size to avoid death-by-fees.

When two NewsCoins about the same story diverge in price by >X%, alert; optionally fire a long/short pair if both are still live.

Auto-claims your accumulated creator fees across all your launched NewsCoins on a schedule. Threshold-gated so you don't burn gas on dust.

Drop a URL or 2-paragraph headline; get a launchable name, ticker, hero quote, and 3-bullet thesis tuned to the platform's `/api/newscoin/suggest` style. For people who launch faster than they write copy.

Walks a first-time launcher through pitch → ticker → curve config → social tease → launch tx. Prevents the "launched at midnight with no one watching" failure mode.

When you bought, you wrote down why ("Coinbase IPO -> alts season"). This skill checks weekly whether the thesis is playing out by re-querying related news + price action and grades it. Forces reflection.

Detects volume spikes >3σ above 24h baseline on coins you watch or hold. Fires fast (sub-minute) so you can react before the trending feed catches up.

Flags suspicious patterns on a coin: rapid buys from <5 wallets, late dump by same wallets, wash-trading footprint. Saves you from being exit liquidity.

Export your full NewsCoin trade history as CSV with USD basis at trade time, ready for Koinly/CoinTracker/your accountant. Settles the "but what was the cost basis" argument.

Weekly DM digest: top 10 NewsCoin creators by earnings, biggest movers, highest-volume coins. One screen instead of clicking through six pages.

A bonding-curve coin's economics shift when it crosses segment transitions. This skill alerts when one of your holdings is approaching a transition so you can pre-position.

Given a coin you hold, finds the 5 most correlated NewsCoins by hourly returns (90d window). Helps you avoid building a "diversified" bag that's actually one bet.

For a coin, fuses on-chain trade flow with IronFeed comment sentiment around the underlying story. Gives you a single bullishness number you can compare across the leaderboard.

Pick a post, get 3 reply drafts in your prior posting voice (analyzed from your last 50 posts). You pick + ship; nothing posts without your tap.

Quote-post a trending post with your contrarian or supporting take, generated from your historical positions on similar topics. Differs from generic reply tools by remembering your stance.

Daily list of 20 accounts you should follow based on your engagement patterns and graph distance — already-engaged-with-you accounts excluded. Lifts followback rate by skipping the obvious.

Each morning, queues 5–10 posts in your niche worth a thoughtful reply, with drafted replies. Builds presence without doomscrolling. Approve in <2 minutes.

Long-form input (essay, transcript, doc) → 5–10 post thread sized to IronFeed character limits, with hook post and CTA. Auto-numbers; auto-pulls quotes.

Given your draft, suggests 3 tags that lift visibility based on current trending volume vs. your follower-tag overlap. Avoids the "everyone uses #crypto so it's noise" trap.

Analyzes your last 90 days of impressions/engagement and outputs the 3 best windows for you (timezone-localized) with confidence intervals. Stops debating "is morning better."

Suggests accounts to mute based on your dwell time + reply patterns (you keep scrolling past or rage-replying). Every keep-mute is a vote for sanity.

Mirror your X posts (via the connected Nitter feed) into IronFeed with attribution and your wallet's badge. Saves typing twice.

Auto-replies to `ironclaw-alerts` posts that mention you (e.g., "@you flagged in proposal #41"), drafting a measured response and flagging the urgent ones for human review.

When a verified poster quotes or cites you, broadcasts a follow-up to your own followers with a "here's what they missed" angle. Turns one mention into a small wave.

Collapses 200+ comments on a post into a 5-bullet recap of where consensus + dispute land, so you can reply to the conversation, not the noise.

Score a poster 0–100 on track-record reliability across their last 12 months: NewsCoin call hit rate, governance votes vs. outcomes, follower retention.

Daily list of new accounts gaining traction in tags you specify (engagement growth >50% week over week, started <60 days ago). Catches future heavyweights early.

Daily 1-paragraph "what changed in the discourse today" — clusters trending tags into narratives, names them, attaches example posts.

A specific post underperformed; this skill compares it to your last 20 hits and explains why (timing, hook strength, tag mix, follower decay).

Drop a topic; get a poll question plus 4 distinct, defensible options that drive thoughtful votes. Avoids the lazy "yes/no/maybe/idk" pattern.

Finds high-engagement posts (top 5% by impressions) in your niche with <10 replies — the open mic moments. Lists 5 a day, with drafted opening replies.

Classifies inbound DMs into urgent / business / info / spam buckets and routes them. Drafts replies for "info" requests so you only handle the urgent ones live.

Stronger than out-of-office: drafts a contextual reply for any inbound DM (knows the thread history), holds it for your one-tap approval before send.

Plain "I'm out, back on X" auto-reply with optional escalation rule (urgent keywords still get pinged via TG). Cheaper, simpler than triager — for vacations only.

Once a day, a 3-bullet recap of unread threads ("Alice asked about X", "Group ABC pivoting strategy"). Run after coffee instead of opening 14 tabs.

First DM from a new peer? Auto-prompts verification flow ("Confirm fingerprint: ABC123 with them out-of-band"). Flags any peer whose key changes after verification.

Pulls all $TICKERS, contract addresses, links, and named wallets out of a DM thread into a clean table. Handy when a friend dumps research at you.

Silently moves low-trust senders (new wallets, no followers, scam-list match) to a quarantine bucket. Doesn't auto-reply (replying confirms aliveness).

If you haven't replied to an opened thread in N hours, nudges you with a draft reply. Specifically for threads you marked "important", not all unreads.

Daily/weekly recap for a chosen group DM (plaintext groups only). Names the threads, top opinions, and any decisions.

Scans your last 7 days of DMs and outputs a TODO list with source links. Nothing fancier — just "you said you'd send Alice the deck. Still need to."

Inside a thread, polls both sides for availability windows, picks an overlap, drops the calendar invite. Stops the "what about Tuesday" tennis match.

Translates inbound foreign-language DMs into your preferred language inline; preserves the encrypted source so the recipient still owns the plaintext.

Walks you through rotating your DM keypair safely (back up old key, distribute new key per active peer, mark stale). Most users won't otherwise.

Before send, scans your draft for accidentally-pasted private keys, seed phrases, API keys, screenshots with EXIF leaks. One late-night save can repay a year's subscription.

Reviews your draft DM before send; flags if tone reads passive-aggressive, accusatory, or off-brand for the relationship type. One-tap rewrite.

Simulate the actual fill price for a given size against a chosen pool's depth, factoring fees and route splits. Stops the "I expected 0.4% slippage and got 6%" surprise.

Mirror trades from a wallet you trust, with a per-trade cap and a daily ceiling. Skips trades into tokens flagged by `risk-token-scorer`.

For any wallet, computes hit rate, ROI per trade, max drawdown, average hold, win/loss ratio over the last 90/180/365 days. Backs claims of "I'm up 4x" with math.

Watches your open positions for TP/SL hits and pings via DM and TG. Different from native price alerts — knows about your entry, not just price.

For an LP position, projects expected IL across price scenarios and against just-holding. Saves the "why am I down vs spot" head-scratch.

Sweeps tiny token balances (under N NEAR equivalent) into NEAR or USDC in a single batched route. Cleans up the long tail.

Finds momentary cross-chain price gaps for the same asset (USDT, ETH, NEAR) factoring 1-click bridge fees. Lists actionable gaps, doesn't fire by default.

Compares routes across direct DEX, NEAR Intents 1-click, and same-chain hops; picks the cheapest considering bridge fees and slippage.

FIFO/LIFO lot accounting across all your tracked positions, exported as CSV with realized/unrealized PnL. Includes NewsCoin trades.

Of your last 100 trades, what % were rugs / honeypots / softrugs vs runners? Honest mirror of your filter quality. Pairs well with `risk-honeypot-detector` so you can A/B yourself.

Grades your current open portfolio A–F on risk: concentration, time-in-market mix, scam-list exposure, illiquid bag tail. One letter and a 2-bullet why.

Each entry, log a one-paragraph thesis. On exit, the skill grades the thesis vs outcome and adds it to your trader journal. Compounds: you stop repeating the same mistake.

Trailing stop-loss on a NewsCoin or on-chain position; trails N% below the high-water mark. Fires sell tx when breached.

Alerts when a token in your watchlist has an intraday move >Xσ vs its 30d realized vol. Catches breakouts and capitulations before the chat does.

Radar of small caps (<$10M MC) showing >50% volume + holder growth in last 24h on chains you trade. Differs from generic trending lists by filtering for actionable size.

Watch a list of whale wallets; alert when ≥N of them buy the same token within a window. Stops "who is this whale, did anyone else front-run him" guessing.

Weekly digest: how much you paid in fees, broken down by chain/protocol/skill, and where to cut. Most people are paying 3x what they think.

Checks if your last N trades show sandwich-attack patterns (front-run + back-run by same wallet). Quantifies the bleed and recommends private-RPC routes.

Maintains a multi-token portfolio at target weights; rebalances when drift exceeds threshold. Different from NewsCoin rebalancer — works across chains and includes spot tokens.

Schedules a non-urgent swap for the historically cheapest gas window of the next 24h on the chosen chain. Saves gas, not edge.

TL;DR every active proposal in 5 bullets: what it changes, who proposed, voting deadline, financial impact, conflicts. Linked to the source proposal id.

Walks you through 5 questions to capture your values, then recommends a vote on each active proposal with a rationale. Updates as your stated values change.

Given your past votes (or your stated values), finds the 3 delegates whose vote pattern most aligns. Stops "who do I delegate to" being a 4-hour Telegram quest.

For mission proposals (the agent-economy escrow + payout pattern), spells out who pays, who claims, payout schedule, and risk to treasury if claim fails.

Alerts when a proposal you've flagged is short of quorum within 24h of close. Optionally auto-casts your pre-committed vote.

Flags when two pending proposals contradict each other ("Proposal A sets fee 10%, Proposal B sets fee 12% on same surface"). Surfaces governance hygiene issues early.

Export your full vote history as CSV with rationales (if logged via vote-recommender). Useful for delegate accountability or just reflecting on past stances.

Turns a paragraph of intent into a well-structured proposal (title, summary, rationale, expected impact, voting question). Doesn't post; you review and submit.

For agent-prompt change proposals, shows the actual diff vs current on-chain prompt — character-level, not just summary. Voters know what they're approving.

Reconciles every treasury proposal that passed against actual on-chain spend. Surfaces unspent allocations, overspends, and proposers with bad track record.

Public ranking of voters by % of votes that aligned with the eventual outcome. Reputational, not financial — but creates a long-term incentive to vote thoughtfully.

Recommends a stake-lock duration that maximizes voting power vs. opportunity cost of locked NEAR, given upcoming proposals you care about.

Find historic proposals semantically similar to the one in front of you and how they were ultimately resolved. Stops the community relitigating settled questions.

Vanguard NFT holders earn revshare drops; this skill claims them on a schedule once accumulated value passes threshold. Stops gas burn on dust.

Watches room chat for rule-breaking content (slurs, scam links, stake-evasion), suggests mute/kick to host with one-tap actions. Doesn't act unilaterally.

On room close, generates a 5-bullet recap post with the recording link, top alpha calls, and standout speakers, posted to IronFeed.

When a speaker calls an alpha (e.g., "I'm long $NEAR here"), fact-checks claims against on-chain data in <5s and posts a green/yellow/red flag to chat.

Stake-aware speaker queue: hand-raises sorted by amount of stake the listener has locked, so high-conviction questions go first. Built for high-value alpha rooms.

Records every alpha call in a room (with vote counts) and tracks 24h/7d/30d performance. Builds an alpha track record for that room and each speaker.

Collects listener questions throughout a room, dedupes/clusters, hands the host a top-5 list at chosen marks. Avoids host scrolling 200 raised hands mid-flow.

Turns the egress mp4 into chapters + transcript bullets + auto-generated quote cards. The "I missed the room, give me the gold" surface.

Runs the run-sheet for a multi-host room (intro, segment timing, handoff cues, sponsor reads) and pings hosts in DM at the right beats.

30 minutes before going live, drafts and posts a teaser to your IronFeed with topic, speakers, and stake floor. Stops the "we're live, but no one knew" failure.

Distributes small NEAR drops to listeners who stayed >X minutes. Tiny, but it boosts retention measurably and creates a culture of showing up.

Hosts get a post-room score for the room: question count, alpha-call vote ratio, retention curve, top-talker ratio. Stops them flying blind on whether the room landed.

Weekly cross-room leaderboard of best alpha calls (24h hit), with speaker handles and rooms. Rewards the room ecosystem's actual signal-makers.

Recommends a join-stake based on topic value and historic participation in similar rooms. Stops hosts under-pricing alpha rooms or scaring listeners away with too-high stakes.

Bundles recording mp4 + recap + chat highlights + alpha-call leaderboard into a single shareable post. The "did we have a room?" deliverable for sponsors.

End-to-end first-day check: token allowances, scam-list exposure, key custody (custodial vs hardware), DM verification status. Single A–F grade plus the 3 things to fix today.

Step-by-step "your first NewsCoin": pick from trending, explain the curve, walk through the buy at safe size, log the thesis. Holds your hand without being condescending.

Picks a recommended kit based on user's stated goals (degen / creator / governance nerd / DM-heavy), installs it, configures presets, tests one end-to-end run.

Walks through the unhappy paths: lost seed (use NEAR account recovery), partial key compromise (rotate DM key, revoke allowances), wallet drained (file report, freeze).

Lists token allowances ranked by risk (unlimited > large > stale > small), revokes selected ones in a batched tx. Saves you from the next one-click drain.

Compromised? Walks through emergency freeze: pause all running automations, revoke allowances, rotate DM key, post a public notice if relevant. Coordinates the chaos.

Import positions from another wallet (yours, not the user's other random wallet) so the trading terminal and skills can reason about your full bag. Read-only.

Explains Pro perks (higher AI budget, badge, themes), checks $IRONCLAW balance, locks the right amount for the chosen tier, confirms activation.

New user with empty feed? Picks 50 first follows from interests + persona heuristics. After 7 days, suggests prunes for accounts that didn't earn engagement.

Shows the user what they need to qualify for Vanguard NFT (history, holdings, governance participation), with a tracker. Most are 80% there and don't realize.

Walks through linking the Telegram bot: mints code, points user at t.me/<bot>?start=<code>, confirms link, configures notifications.

Gives the user's wallet a hygiene tier (Bronze/Silver/Gold/Platinum) with the specific actions needed to advance. Different from `onboard-security-audit` — about positive growth, not just risk.

Sub-200-word morning brief on your watched topics with markets, governance deadlines, mentions. Improved over the existing built-in: persistent topics + tone choice.

Sunday-evening one-pager: PnL, governance votes, social engagement, missions claimed/completed, top wins, top regrets. Closes the week deliberately.

Outputs an ICS feed of your AZUKA events: live rooms you said you'd attend, governance deadlines, mission deadlines, scheduled automations.

When you post on X, mirror to IronFeed automatically (with attribution and the "originally on X" badge). Different from `feed-cross-poster` by being one-way + automatic.

Natural-language reminders: "remind me when ABC hits 0.5 NEAR", "remind me about Alice's proposal Tuesday", "remind me to claim creator fees Friday". Routes to TG/DM.

Save posts and links to a queue; daily summarizes the queue and pings you with the 3 most important ones to read first. Beats bookmarking 80 things you'll never see again.

Single screen of every watched token + watched poster, color-coded for movement. Renders in a DM/TG card so you don't open a 6-tab watchlist app.

Each week, surfaces bookmarks you saved but never read or acted on. Gentle nudge plus a 1-line "still relevant?" tag.

Before a 1:1 (DM or scheduled call), one-page prep on the counterparty: last 10 posts, current holdings of relevance, mutual contacts, recent governance votes.

Record weekly/monthly goals ("ship a skill", "vote on every proposal", "post 5x"); skill grades end-of-period from on-chain + IronFeed activity.

End-of-day checklist: TP hits to acknowledge, gov votes due in 24h, unread DMs flagged urgent, automations that errored. Close the loop.

Tracks daily streaks for posting / voting / trading and celebrates milestones in feed/TG. Light gamification proven to lift retention 30–40%.

Compares 1-click bridge vs. swap-then-bridge vs. bridge-then-swap routes for the same A→B and surfaces the cheapest, including dest-chain claim gas. Different from `trading-route-optimizer` by including chain hops.

Estimates worst-case slippage on a chosen bridge route for a given size, including stablecoin de-peg risk on small rails.

Projects total gas across the full bridge: source send + bridge fee + dest claim + dest swap if applicable. Gives you a single USD number you can decide on.

Watches a bridge deposit address until COMPLETE/REFUNDED and pings you. Stops the obsessive every-30-seconds refresh after a bridge.

When a bridge stalls (>30 min in PROCESSING), diagnoses likely cause from logs + suggests next step (refund poll, contact support, retry). Specifically for the 1-click long-tail.

When direct A→B isn't supported, plans a multi-hop A→C→B with split-execution to minimize peg-risk window. Useful for thin pairs.

Monthly digest of your bridge fees, broken down by route, with a "you would have saved $X if you'd used route Y" callout.

Given a source asset, lists every reachable destination asset with sample quotes. Stops "can I bridge X to Y" Discord questions.

Tags every bridge event in your history (transfer / swap / cost-basis-event) with explanations, ready to drop into a tax tool. Bridges are the #1 thing tax tools mis-categorize.

Before sending funds, sanity-checks the deposit address against the quote (amount, asset, expiry, signer). Catches the "wait, that's not the right chain" mistake.

Pre-trade risk score for a token: liquidity depth, top-10 holder concentration, contract verification, age, scam-list match. One letter + 3 facts.

Checks a URL or post against the scam list and a learned phishing-pattern model; returns a verdict and a draft warning to send to the recipient.

Lists wallet allowances ranked by risk (unlimited > large > stale > small) with "kill" links. Different from `onboard-allowance-pruner` by being an ongoing monthly check, not one-shot.

Checks the user's wallet for drain-pattern indicators (recent allowance set + new token added + outgoing native transfer to fresh address) and pings urgent if found.

Trust score for a poster across IronFeed history: account age, mute-rate, scam-flag history, vote alignment, NewsCoin call hit rate. Different from `feed-poster-credibility` — emphasizes safety, not signal.

For a given room, scores how reliable past alpha calls have been (24h hit rate, downside-of-misses). Helps you decide if a join-stake is worth it.

Before installing a third-party kit, audits manifest hash, included skills' verified flag, revenue-split sanity, and any required webhook destinations.

Before installing a third-party HTTP skill, checks verified flag, callback domain reputation, install count vs. age, author wallet history.

Before swapping with / DMing / sending NEAR to an address, returns a reputation card: account age, scam flags, related wallets, public posts.

Scans your last 30 days of public posts for accidentally-leaked sensitive info: full wallet addresses tied to amounts, screenshots with text reveals, API keys.

Before claiming a mission, verifies escrow is funded, payout terms unchanged, claimant can actually meet acceptance criteria. Stops "I did the work, where's the money."

Pre-trade simulation: tries a buy + immediate sell on a fork, flags tokens that allow buy but block sell. Different from token-scorer — actively probes.

Classifies a NewsCoin's rugpull risk: creator wallet history, supply structure, early-buyer concentration, segment behavior. Specific to the NewsCoin curve, not a generic ERC-20 model.

Walks a heavy user through promoting their wallet to a multisig (2-of-3 with a hardware key + recovery key + warm key). Coordinates the keypair generation, invite flow, deploy.

Auto-replies to certain TG DMs while you're heads-down (custom rules per sender), drafts urgent ones for your one-tap approval.

Daily summary of a chosen Telegram group, posted to your TG DM. The "I'm in 14 groups but couldn't read three of them today" surface.

Watchlist-aware pump alerts pushed to TG; one-tap "buy 0.5 NEAR" inline keyboard. Already-wired tg watchlist endpoints; missing piece is the inline-action button.

Push notifications when proposals you care about move (status change, quorum reached, vote about to close), with one-tap vote inline keyboard.

`/balance` returns a rich card with native + token balances, 24h PnL, top 3 movers. More info per glance than the default.

`/quick <amount> <ticker>` initiates a swap with confirm-once flow. Stops the 8-message "swap, confirm, where, how much" tennis match.

Alert when your @ is mentioned in a watched TG group (without joining DM bridge). One-line ping with link to the message.

PIN-gated cold-storage helper. `/freeze` pauses all skill-driven trades; `/withdraw <amount>` requires a separate PIN before signing. Stops fat-finger sends from /quick-trade.

Silences non-urgent TG notifications between configured hours (only urgent-tagged messages and DMs from a whitelist break through). Sleep matters.

`/wallet <name>` switches the active wallet for trading commands. Existing `/api/tg/settings` supports this, this skill makes it discoverable + adds confirmation.

`/portfolio` returns a detailed card: NewsCoin holdings, on-chain positions, gov stake, Vanguard NFT, mission earnings, and a 1-line summary.

Inside TG, walks a new user through linking, first buy, first vote, first follow. Different from `onboard-tg-link-helper` — this is full AZUKA onboarding via TG, not just the link step.

`/freeze` immediately pauses all your active skill automations and revokes a sliding window of allowances. Designed for "I clicked something, am I cooked" moments.

Creator broadcast: send a message to all groups you admin (with rate-limit + rule-of-three confirmation). Avoids "I have to copy-paste this 11 times."

Break down skill sales, NewsCoin creator fees, room revenue, kit royalties; per-skill, per-buyer, daily/weekly chart. Stops the "I made $X this month from somewhere" mystery.

Ranks your top fans across DMs, replies, room participation, skill installs. Shows the 20 names you should DM personally this week.

Plans a 7-day content schedule (mix of posts/threads/quote-takes/poll), drafts each, queues at best times from `feed-best-posting-time`.

Split incoming royalties (from skill sales or NewsCoin fees) to up to 5 wallets at fixed bps. Useful for collabs.

Auto-airdrops a small NEAR or token reward to your top N fans on a cadence. Cements loyalty cheaply.

Looks at install rate, churn, refund signal, and competitor prices in your category; recommends a price change with confidence interval.

When you launch a new skill, generates 3 launch posts (hook + carousel + thread) tuned to your follower base, posts on schedule.

Before launching a paid skill: pricing reviewed, manifest fields complete, hero image present, demo video link, support DM link, refund policy. Catches the small misses.

Side-by-side benchmark of your skill vs the top 5 in its category: install count, retention, price, review sentiment. Different from `creator-skill-pricing-coach` — qualitative comparison, not pricing-only.

Finds your top 10 historical posts (by impressions/engagement) and proposes 3 fresh-format remixes each (thread, video script, room topic).

Sets up a NEAR tip jar tied to your profile and posts; auto-acknowledges with a thanks DM and shoutout in your weekly recap. Different from `creator-fan-dm-broadcast` — passive incoming, not outgoing.

DM your subscribers (opt-in list) with a release note when you ship a new skill, post, or coin. Rate-limited; refusing inbound DMs after broadcast (no spam).

Semantic search across IronFeed posts (your follows + niche tags). Different from native trending — finds the post you remember reading 3 weeks ago.

Alert when a tag enters or exits the trending top-N (configurable). Catches the moment a niche becomes mainstream.

Suggested follows updated weekly based on accounts your engaged-with people engage with (graph-2 distance) but you don't yet follow. Beats "rec by tag" by 3x.

Surfaces emerging niches (clusters of <100-follower accounts gaining 50%/wk engagement on a shared tag). For people who want to be 6 months early.

Matches you with open missions whose required skills + history match yours. Different from generic mission listings — ranks by your fit, not recency.

Right now, what live room aligns with your interests? Real-time, not "what's scheduled tomorrow." Useful only because rooms are emergent.

Recommends skills you'd benefit from based on your wallet activity, gaps detected in onboarding audit, peer install patterns. Smart upsell, but earns trust by being genuinely useful.

Recommends a kit (or kit upgrade) based on which skills you've installed, gaps, and installed-by-similar-users patterns.

Find emerging creators in your niche to back early — defined as: <500 followers, top-quartile engagement-per-follower, posting >3x/week, scam-flag-free.

Given a NewsCoin you hold, finds 5 related (correlated narrative or holder overlap) NewsCoins you should consider — for diversification or theme amplification.

Inside a DM negotiation, proposes tiered counteroffers (e.g., 10/20/30 NEAR with different terms each) to nudge convergence. Stops the back-and-forth-on-price stall.

When you're negotiating the same thing with 3 buyers, this skill keeps state per thread and surfaces the highest-bid + most-willing combination. The "best deal" picker.

Drafts mission/escrow terms (acceptance criteria, payout schedule, dispute path) from a 1-paragraph deal description, ready to post as a mission.

Side-by-side comparison of 2–3 offers across price, payout schedule, terms strictness, counterparty trust. Beats screenshotting offers next to each other in Notes.

Given an OTC need (e.g., "I want to sell 5,000 USDC in NEAR off-market"), finds counterparties from public history of similar deals + scoring. Coordinates intro DMs.

When a deal stalls or breaks, proposes 3 paths (compromise, escalate to mission arbiter, walk away) with rationale and predicted outcomes from similar past deals.

Proposes a multi-currency payment plan (e.g., "60% USDC now / 40% NEAR vested over 90d") for a deal of size N. Useful for cross-chain settlements.

Reads the counterparty's last 3 messages and grades how flexible they actually are (semantic flags: "I can probably", "this is final", silence patterns). Stops you wasting the day pushing on a sealed door.

Drafts a post in your prior posting voice given a topic + tone. Voice profile is recomputed weekly from your last 50 posts.

Generates 3 image prompts for a given post — for use in Midjourney/SDXL/etc. (no image generation in-platform yet). Prompts capture mood, composition, palette.

Generates a launch pitch (name, ticker, hero quote, 3-bullet thesis) specifically for the NewsCoin curve. Different from `newscoin-pitch-writer` — full launch package, not just naming.

Drop a URL (article, paper, post); get a 5–10 post thread that respects character limits, with a hook post and a CTA. Different from `feed-thread-composer` — single URL input, not raw text.

Pulls 3–5 quotable lines from a long source (essay, transcript, podcast notes) for use in posts/cards. Tuned for "would someone screenshot this" punchiness.

Generates 5 punchy taglines for a project/skill/coin. Specifically tuned to crypto-native attention spans (under 8 words, contains a verb).

Rewrites your IronFeed bio for clarity + niche signal, using your post history to anchor authenticity. A/B tests two variants over 7 days against follow rate.