agentsclimarketplace

Technical scan

Skill NatsuFox/A-Stockit/skills/astockit/technical-scan

A-Stockit —— 面向 Agent 框架的 A 股量化分析技能库,提供多样化市场操作,无需配置独立 trading bot

Install
npx -y skills add NatsuFox/A-Stockit --skill technical-scan

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Run a technical-only scan over one symbol. Use when user explicitly wants trend, levels, volume, volatility, or chart-pattern context without broader thesis, sizing, or execution claims.

SKILL.md

5.2 KB, as published. Nobody here has run it

Technical Scan

Run a technical-only scan for: $ARGUMENTS

Overview

  • Implementation status: workflow-only
  • Current backing path: compose market-analyze plus any needed direct data or saved-artifact inspection
  • Primary purpose: isolate the technical lens from broader narrative, portfolio, or execution layers
  • Research layer: descriptive technical commentary only
  • Local limitation: no dedicated standalone pattern engine or persisted technical-scan artifact exists yet

Use When

  • The user explicitly wants technical context rather than a full report.
  • The caller wants to discuss trend structure, moving-average alignment, support and resistance, volume behavior, or chart language separately.
  • The user wants a technical checkpoint before moving to decision or strategy layers.
  • The user wants technical invalidation or monitoring conditions without broader thesis framing.

Do Not Use When

  • The user wants a full one-symbol report. Use market-brief.
  • The user wants a direct action, quantity, or position recommendation. Use decision-support.
  • The user wants structured execution zones and checklist output. Use strategy-design.
  • The user wants broader thesis, catalysts, or disconfirming evidence beyond technicals. Use analysis.
  • The user wants only the bounded scored state summary. Use market-analyze.

Inputs

  • One stock symbol.
  • Optional local market-data source when the host framework chooses to supply it.
  • Optional neighboring context from market-analyze, market-data, or a persisted artifact.
  • Optional timeframe or horizon if the host framework supports it. If absent, do not pretend a separate timeframe analysis exists.

Execution

Step 1: Anchor the symbol and technical basis

Prefer an explicit symbol plus either:

  • a fresh market-analyze result
  • a named market-data or stock-data artifact
  • a directly inspected local dataset

If the symbol or evidence basis is reused from session context, say so.

Step 2: Reuse the descriptive foundation honestly

Use market-analyze for the core state summary:

  • score and bias
  • trend and regime
  • support and resistance
  • notes and risk flags
  • snapshot-level volume and ATR context

Do not claim that market-analyze directly provides a full chart-pattern engine or every oscillator under the sun. If extra technical color comes from direct data inspection rather than the state summary, label that boundary explicitly.

Step 3: Add technical interpretation only where evidence supports it

If the available data or artifact supports it, the agent may extend the scan with:

  • pattern language such as consolidation, breakout attempt, pullback, rejection, or exhaustion
  • moving-average alignment commentary
  • volatility compression or expansion commentary
  • volume-confirmation or volume-divergence commentary
  • exact technical invalidation or monitoring levels

These are agent-composed interpretations, not guaranteed local signal-engine outputs.

Step 4: Deliver a bounded technical note

Keep the answer technical-only and separate:

  • what the snapshot explicitly reports
  • what direct data inspection adds
  • what is heuristic chart interpretation

Output Contract

  • Expected result: a readable technical-only scan rather than a persisted standalone artifact.
  • Caller-facing delivery standard:
    • state the technical basis used, for example fresh market-analyze, direct data inspection, or a saved artifact path
    • keep the answer technical-only and avoid widening into fundamentals, portfolio action, or execution planning unless the user explicitly changes the request
    • include technical invalidation or monitoring conditions when the evidence supports them
    • label pattern language such as breakout, base, exhaustion, pullback, or rejection as heuristic unless the evidence is explicit
  • Local non-guarantees:
    • no dedicated standalone pattern-engine contract
    • no persisted technical-scan artifact by default
    • no promise that every technical claim is directly computed by a local runner

Failure Handling

  • If the symbol cannot be resolved, return readable guidance instead of guessing.
  • If the technical basis is stale, partial, or insufficient, say what is missing rather than fabricating pattern claims.
  • If the user’s request drifts into action or execution planning, route to decision-support or strategy-design instead of silently widening scope.

Key Rules

  • Keep the scope narrow and technical.
  • Do not silently drift into broader narrative analysis.
  • Use market-analyze as the descriptive anchor unless the caller already provides a better technical basis.
  • Be explicit when a conclusion is chart language rather than a directly computed local signal.
  • Technical risk/reward commentary may describe the setup, but should not become a trading instruction.

Composition

  • Usually follows market-data or market-analyze.
  • Often precedes decision-support, strategy-design, or analysis.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.